AI updates, September 2026

Z.ai has raised the input price for GLM 5.2 from $0.432 to $0.624 per 1M tokens, as listed on OpenRouter's models endpoint. Output pricing is unchanged at $3.99 per 1M tokens. The increase applies only to the prompt side of the call.

The OpenAI API pricing documentation sits inside a docs structure that now includes a "Using GPT-6" quickstart, a hosted Agents API with sandboxes and sessions, and separate ChatGPT surfaces for Plugins, Commerce and Ads. No specific rates are stated in the material we captured, so treat this as a structural change to how OpenAI documents models and costs rather than a confirmed rate change.

MoonshotAI has repriced Kimi Latest on OpenRouter. Input tokens fell from $0.3177 to $0.29 per 1M, while output tokens rose from $7.9431 to $9 per 1M. The net effect depends heavily on how much text your workloads generate.

Qwen3.8 27B output pricing has dropped from $4.35 to $2.99 per 1M tokens on OpenRouter. Input pricing is unchanged at $0.0249 per 1M tokens. The change affects output-heavy workloads only.

The OpenAI API pricing page sits inside a documentation set that exposes Markdown versions of any page by appending .md to the URL, alongside an llms.txt index of the full documentation. That means pricing, model catalogue and cost optimisation pages are served in a form built for machine retrieval rather than only as rendered HTML.

DeepSeek V4 Flash 0423 input pricing dropped from $0.14 to $0.08176 per 1M tokens, and output from $0.28 to $0.16352 per 1M tokens, as listed on OpenRouter. Both rates fell by the same proportion, so the input to output ratio is unchanged.

Z.ai has raised the input price for GLM 5.2 from $0.25 to $0.41 per 1M tokens, as listed on OpenRouter. Output pricing is unchanged at $3.99 per 1M tokens. The change affects input-heavy workloads most.

DeepSeek V4 Flash Latest output pricing dropped from $1.28 to $0.13068 per 1M tokens, roughly a tenfold reduction. Input pricing moved only marginally, from $0.01 to $0.0099 per 1M tokens. The change is listed on OpenRouter's models endpoint.

DeepSeek V4.1 Flash input pricing has dropped from $0.3 to $0.0198 per 1M tokens, a cut of roughly 93%. Output pricing fell from $1.2 to $0.396 per 1M tokens, around a 67% cut. The new rates are listed on OpenRouter's models endpoint.

DeepSeek Flash Latest output pricing on OpenRouter dropped from $0.6 to $0.396 per 1M tokens. Input pricing moved only marginally, from $0.02 to $0.0198 per 1M tokens. The change is almost entirely on the output side.

DeepSeek V4 Flash 0731 input pricing fell from $0.018 to $0.01 per 1M tokens. Output pricing moved the other way, from $0.32 to $1.28 per 1M tokens, a four-fold rise. The model is now much cheaper to feed and considerably more expensive to generate from.

DeepSeek V4 Pro 0813 input pricing dropped from $1.32 to $0.66 per 1M tokens. Output dropped from $3.96 to $1.98 per 1M tokens. Both rates are now half what they were, per OpenRouter's model listing.

Z.ai has raised the input price for GLM 5.2 from $0.1449 to $0.2087 per 1M tokens, as listed on OpenRouter. Output pricing is unchanged at $3.99 per 1M tokens. The increase applies only to the prompt side of each call.

DeepSeek V4 Pro 0813 input pricing has risen from $0.4 to $1.32 per 1M tokens, according to OpenRouter's model listing. Output pricing has moved from $3.49 to $3.96 per 1M tokens. The input change is the larger shift, at roughly three times the previous rate.

Qwen3.8 27B pricing on OpenRouter has moved in both directions. Input went from $0.0249 to $0.42 per 1M tokens, while output dropped from $4.35 to $3 per 1M tokens.

Z.ai has reduced the output price for GLM Latest from $4 per 1M tokens to $1.50 per 1M tokens, as listed on OpenRouter. Input pricing is unchanged at $0.12 per 1M tokens. The change affects output-heavy workloads most.

Google's Gemini Developer API pricing page now lists two price points for Gemini 3.8 Flash, 3.7 Flash and 3.6 Flash: a rate up to 31 December 2026, and a rate double that from 1 January 2027. Standard input goes from $0.75 to $1.50 per million tokens and output, including thinking tokens, from $3.75 to $7.50, with the same doubling applied to Batch, Flex, Priority and context caching. Grounding with Google Search and Google Maps stays at 5,000 free requests a month across Gemini 3.x models, then $14 per 1,000 requests.

Z.ai has reduced the input price for GLM 5.2 from $0.2339 to $0.1449 per 1M tokens, as listed on OpenRouter. Output pricing is unchanged at $3.99 per 1M tokens. The change makes prompt-heavy work on this model meaningfully cheaper, while generation costs stay where they were.

Z.ai has repriced GLM Latest on OpenRouter. Input has fallen from $0.1428 to $0.12 per 1M tokens, while output has risen from $2.992 to $4 per 1M tokens. The cheaper input is offset by a materially more expensive output rate.

MoonshotAI has reduced pricing for Kimi Latest on OpenRouter. Input has dropped from $0.29 to $0.1941 per 1M tokens, and output from $9 to $7.3625 per 1M tokens.

Qwen3 30B A3B input pricing moved from $0.13 to $0.12 per 1M tokens. Output pricing moved from $0.52 to $0.50 per 1M tokens. Both changes are small reductions listed on OpenRouter's model endpoint.

Z.ai's GLM 5.2 input pricing moved from $0.41 to $0.432 per 1M tokens on OpenRouter. Output pricing is unchanged at $3.99 per 1M tokens. The change is small and affects only the input side.

Tencent's Hy4 preview has dropped in price on OpenRouter. Input moved from $0.834 to $0.7506 per 1M tokens, and output from $2.501 to $2.2509 per 1M tokens, a 10% reduction on both sides.

DeepSeek has trimmed pricing on DeepSeek V4 Flash 0423, as listed on OpenRouter. Input moves from $0.08176 to $0.07854 per 1M tokens, and output from $0.16352 to $0.15708 per 1M tokens. Both are small reductions of roughly 4 per cent.

Kimi Latest from MoonshotAI has dropped its listed rates on OpenRouter. Input moved from $0.3654 to $0.3177 per 1M tokens, and output from $9.1343 to $7.9431 per 1M tokens. Output remains the dominant cost, at roughly 25 times the input rate.

DeepSeek Pro Latest input pricing has dropped from $0.134 to $0.123 per 1M tokens on OpenRouter. Output pricing is unchanged at $3.5 per 1M tokens. The cut is small and applies only to the input side.

Kimi Latest input pricing has moved from $0.4 to $0.3654 per 1M tokens on OpenRouter. Output pricing has gone the other way, from $9 to $9.1343 per 1M tokens. The net effect depends entirely on how output-heavy your workload is.

Z.ai's GLM Latest input price moved from $0.12 to $0.13 per 1M tokens, as listed on OpenRouter. Output pricing is unchanged at $4 per 1M tokens. The shift is small and affects input-heavy workloads only.

Z.ai's GLM Flash Latest has dropped its output price from $0.30 to $0.2475 per 1M tokens, as listed on OpenRouter. Input pricing is unchanged at $0.02 per 1M tokens. The change makes long-output work on the model cheaper, with no movement on the prompt side.

DeepSeek Pro Latest input pricing fell from $0.2 to $0.134 per 1M tokens on OpenRouter. Output pricing is unchanged at $3.5 per 1M tokens. The cut applies to input only, so the gap between reading and generating widens.

Kimi K2.7 Code input pricing moved from $0.6562 to $0.6712 per 1M tokens, and output from $3.3 to $3.35 per 1M tokens, per OpenRouter's model listing. Both changes are small upward adjustments rather than a repositioning of the model.

Z.ai has raised the input price for GLM 5.2 from $0.2087 to $0.25 per 1M tokens, as listed on OpenRouter. Output pricing is unchanged at $3.99 per 1M tokens. The change affects input costs only, so prompt-heavy workloads feel it most.

DeepSeek V4 Flash Latest input pricing has moved from $0.012 to $0.01 per 1M tokens. Output pricing has gone the other way, from $1.25 to $1.28 per 1M tokens. Pricing is listed via the OpenRouter models API.

DeepSeek Pro Latest input pricing has moved from $0.15 to $0.2 per 1M tokens, as listed on OpenRouter's models endpoint. Output pricing is unchanged at $3.5 per 1M tokens. The increase applies to input only, so the gap between input and output costs narrows slightly.

Meta's Muse Glimmer 30B has moved from $0.3 to $0.35 per 1M input tokens, and from $1.2 to $1.5 per 1M output tokens, per OpenRouter's model listing. Output pricing takes the larger increase of the two.

The OpenAI API pricing page under platform.openai.com now sits inside a reorganised documentation structure, with pricing listed under the model catalog alongside model selection. The captured material shows navigation and page structure only, with no per-token rates or model rates included. Nothing in the source confirms a change to what anything costs.

Google's Gemma 4 26B A4B has dropped in price on OpenRouter. Input moved from $0.09 to $0.0765 per 1M tokens, and output from $0.3 to $0.255 per 1M tokens. That is a 15% reduction on both sides.

DeepSeek V4 Pro 0813 input pricing moved from $0.3942 to $0.4 per 1M tokens on OpenRouter. Output pricing is unchanged at $3.49 per 1M tokens. The change is a rounding-scale adjustment rather than a shift in the model's cost profile.

Qwen3.8 27B output pricing has moved from $4.4 to $4.35 per 1M tokens on OpenRouter. Input pricing is unchanged at $0.0249 per 1M tokens. The gap between input and output cost remains very wide.

Pricing | OpenAI API For the complete documentation index, see llms.txt . Markdown versions of documentation pages are available by appending .md to the page URL. ChatGPT Start searching API Dashboard Try ChatGPT Home API Overview Get started with the OpenAI API Models Explore models and compare cap

Name: GPT-6.1 Sol Provider: OpenAI Listed: 2026-09-29 Context window: 1050000 tokens Input: $2 per 1M tokens Output: $10 per 1M tokens GPT-6.1 Sol is an upgrade to GPT-6 Sol from OpenAI, positioned below the flagship GPT-6 Astra in the GPT-6 series. It is suited for agentic coding, computer use, doc

Name: GPT-6.1 Sol Pro Provider: OpenAI Listed: 2026-09-29 Context window: 1050000 tokens Input: $2 per 1M tokens Output: $10 per 1M tokens GPT-6.1 Sol Pro is the same underlying model as [GPT-6.1 Sol](https://openrouter.ai/openai/gpt-6.1-sol), served with `reasoning.mode` set to `pro` for higher-qua

Model: DeepSeek V4 Flash 0423 Input was $0.14 per 1M tokens, now $0.0763 per 1M tokens Output was $0.28 per 1M tokens, now $0.1526 per 1M tokens

Model: DeepSeek V4 Pro 0423 Input was $0.95526 per 1M tokens, now $0.936294 per 1M tokens Output was $1.91052 per 1M tokens, now $1.872588 per 1M tokens

Model: GLM 5.3 Input was $1.4 per 1M tokens, now $0.1849 per 1M tokens Output was $4.4 per 1M tokens, now $4.4 per 1M tokens

Model: gpt-oss-120b Input was $0.15 per 1M tokens, now $0.037 per 1M tokens Output was $0.6 per 1M tokens, now $0.17 per 1M tokens

Model: DeepSeek V3.2 Exp Input was $0.1344 per 1M tokens, now $0.27 per 1M tokens Output was $0.2016 per 1M tokens, now $0.41 per 1M tokens

Model: DeepSeek V4 Flash Latest Input was $0.018 per 1M tokens, now $0.012 per 1M tokens Output was $0.32 per 1M tokens, now $0.5 per 1M tokens

Model: DeepSeek Pro Latest Input was $0.22592 per 1M tokens, now $0.15808 per 1M tokens Output was $1.9584 per 1M tokens, now $1.9584 per 1M tokens

Model: Qwen3 30B A3B Input was $0.12 per 1M tokens, now $0.13 per 1M tokens Output was $0.5 per 1M tokens, now $0.52 per 1M tokens

Model: Qwen3 30B A3B Instruct 2507 Input was $0.1 per 1M tokens, now $0.04815 per 1M tokens Output was $0.3 per 1M tokens, now $0.19305 per 1M tokens

Model: DeepSeek V3.1 Terminus Input was $0.27 per 1M tokens, now $0.3 per 1M tokens Output was $1 per 1M tokens, now $1 per 1M tokens

Model: Kimi Latest Input was $0.85825 per 1M tokens, now $0.85825 per 1M tokens Output was $11.1062 per 1M tokens, now $11.9 per 1M tokens

Model: GLM 5.2 Input was $0.6496 per 1M tokens, now $0.3249 per 1M tokens Output was $2.0416 per 1M tokens, now $3.52 per 1M tokens

Model: DeepSeek V4 Pro 0813 Input was $0.3915 per 1M tokens, now $0.4 per 1M tokens Output was $4.2 per 1M tokens, now $4.2 per 1M tokens

Model: Qwen3.8 27B Input was $0.42 per 1M tokens, now $0.0449 per 1M tokens Output was $3 per 1M tokens, now $4.4 per 1M tokens

Model: GLM Latest Input was $0.15 per 1M tokens, now $0.15 per 1M tokens Output was $3.3 per 1M tokens, now $4 per 1M tokens

Model: GLM Flash Latest Input was $0.045 per 1M tokens, now $0.03 per 1M tokens Output was $0.14 per 1M tokens, now $0.5 per 1M tokens

Model: DeepSeek Flash Latest Input was $0.0449 per 1M tokens, now $0.02 per 1M tokens Output was $0.7 per 1M tokens, now $0.6 per 1M tokens

Tarifs de l'API Gemini Developer | Gemini API | Google AI for Developers Passer au contenu principal / English Deutsch Español – América Latina Français Indonesia Italiano Polski Português – Brasil Shqip Tiếng Việt Türkçe Русский עברית العربيّة فارسی हिंदी বাংলা ภาษาไทย 中文 – 简体 中文 – 繁體 日本語 한국

Pricing - Claude Platform Docs Claude Platform Docs Messages Managed Agents Admin Resources  Best practices Models & pricing  SDKs, CLI, and libraries Claude API skill Release notes  API reference  English   Console Log in   Search Ctrl K Models Models overview Claude Fable 5.1 Claude Opus 5

DeepSeek V4 Pro 0423 input pricing moved from $0.783 to $0.95526 per 1M tokens, and output from $1.566 to $1.91052 per 1M tokens, according to OpenRouter's model listing. Both sides rose by the same proportion, so the input to output ratio is unchanged.

DeepSeek Flash Latest input pricing moved from $0.035 to $0.0449 per 1M tokens on OpenRouter. Output pricing moved from $0.29 to $0.7 per 1M tokens, a larger jump than the input change.

MoonshotAI has cut pricing on Kimi Latest as listed on OpenRouter. Input moves from $0.875 to $0.85825 per 1M tokens, and output from $11.2 to $11.1062 per 1M tokens. Both changes are small, under two percent.

DeepSeek has lowered the input price on DeepSeek V4 Flash 0731 from $0.021 to $0.018 per 1M tokens, as listed on OpenRouter. Output pricing is unchanged at $0.32 per 1M tokens. The change is input-side only.

DeepSeek V4 Flash Latest input pricing has moved from $0.019 to $0.018 per 1M tokens on OpenRouter. Output pricing has gone the other way, from $0.3 to $0.32 per 1M tokens. The net effect depends on how output-heavy your workload is.

NVIDIA has lowered pricing for Nemotron 3.5 Lightning on OpenRouter. Input drops from $0.08 to $0.06 per 1M tokens, and output drops from $0.20 to $0.16 per 1M tokens. That is a 25 percent cut on input and 20 percent on output.

Input pricing for DeepSeek V4 Pro 0813 moved from $0.3955 to $0.3915 per 1M tokens on OpenRouter. Output pricing is unchanged at $4.20 per 1M tokens. The change is a fractional adjustment rather than a repositioning of the model.

Model: Grok Latest Input was $1.6 per 1M tokens, now $2 per 1M tokens Output was $4.8 per 1M tokens, now $6 per 1M tokens

Model: GPT-5.6 Sol Pro Input was $2 per 1M tokens, now $4 per 1M tokens Output was $10 per 1M tokens, now $20 per 1M tokens

Model: DeepSeek V4 Flash Latest Input was $0.021 per 1M tokens, now $0.019 per 1M tokens Output was $0.32 per 1M tokens, now $0.3 per 1M tokens

Model: Grok 4.7 Input was $1.6 per 1M tokens, now $2 per 1M tokens Output was $4.8 per 1M tokens, now $6 per 1M tokens

Name: Claude Sonnet 5.5 Provider: Anthropic Listed: 2026-09-28 Context window: 1000000 tokens Input: $2 per 1M tokens Output: $10 per 1M tokens Claude Sonnet 5.5 is Anthropic's Sonnet-class model for well-scoped everyday work, succeeding Claude Sonnet 5 as a direct upgrade. It is especially strong a

Pricing - Claude Platform Docs Claude Platform Docs Messages Managed Agents Admin Resources  Best practices Models & pricing  SDKs, CLI, and libraries Claude API skill Release notes  API reference  English   Console Log in   Search Ctrl K Models Models overview Claude Fable 5.1 Claude Opus 5

Model: DeepSeek V3.2 Exp Input was $0.27 per 1M tokens, now $0.1344 per 1M tokens Output was $0.41 per 1M tokens, now $0.2016 per 1M tokens

Model: Nex-N2.5-Pro Input was free, now $0.075 per 1M tokens Output was free, now $0.25 per 1M tokens

Model: Nex-N2.5-Mini Input was free, now $0.025 per 1M tokens Output was free, now $0.1 per 1M tokens

Çmimet e API-t të Zhvilluesit Gemini | Gemini API | Google AI for Developers Kapërce te përmbajtja kryesore / English Deutsch Español – América Latina Français Indonesia Italiano Polski Português – Brasil Shqip Tiếng Việt Türkçe Русский עברית العربيّة فارسی हिंदी বাংলা ภาษาไทย 中文 – 简体 中文 – 繁體 日本語

Model: Gemma 4 26B A4B Input was $0.0675 per 1M tokens, now $0.09 per 1M tokens Output was $0.225 per 1M tokens, now $0.3 per 1M tokens

Model: Kimi K2.6 Input was $0.95 per 1M tokens, now $0.65 per 1M tokens Output was $4 per 1M tokens, now $3.41 per 1M tokens

Model: DeepSeek V3 0324 Input was $0.25 per 1M tokens, now $0.29 per 1M tokens Output was $1 per 1M tokens, now $1.14 per 1M tokens

Model: DeepSeek V3.2 Input was $0.269 per 1M tokens, now $0.28 per 1M tokens Output was $0.4 per 1M tokens, now $0.42 per 1M tokens

Model: Qwen3.5-35B-A3B Input was $0.3125 per 1M tokens, now $0.1625 per 1M tokens Output was $1.25 per 1M tokens, now $1.3 per 1M tokens

Model: MiniMax M2.7 Input was $0.21 per 1M tokens, now $0.3 per 1M tokens Output was $0.84 per 1M tokens, now $1.2 per 1M tokens

Model: DeepSeek V4 Flash Vision Exp Input was $0.2156 per 1M tokens, now $0.44 per 1M tokens Output was $0.6468 per 1M tokens, now $1.32 per 1M tokens

Model: GLM 5.3 Flash Input was $0.045 per 1M tokens, now $0.15 per 1M tokens Output was $0.14 per 1M tokens, now $0.5 per 1M tokens

Model: GLM 5.1 Input was $0.9646 per 1M tokens, now $1.4 per 1M tokens Output was $3.0316 per 1M tokens, now $4.4 per 1M tokens

Model: DeepSeek V4 Flash 0423 Input was $0.0469 per 1M tokens, now $0.14 per 1M tokens Output was $0.0938 per 1M tokens, now $0.28 per 1M tokens

Model: DeepSeek V4 Pro 0423 Input was $0.348 per 1M tokens, now $0.9483 per 1M tokens Output was $0.696 per 1M tokens, now $1.8966 per 1M tokens

Model: Kimi Latest Input was $1 per 1M tokens, now $0.995 per 1M tokens Output was $9 per 1M tokens, now $9 per 1M tokens

Model: DeepSeek V4 Pro 0813 Input was $0.29304 per 1M tokens, now $0.24298 per 1M tokens Output was $3.5 per 1M tokens, now $3.5 per 1M tokens

Model: GLM 5.3 Input was $0.3556 per 1M tokens, now $0.3556 per 1M tokens Output was $1.1176 per 1M tokens, now $2.574 per 1M tokens

Model: GLM Latest Input was $0.3556 per 1M tokens, now $0.3556 per 1M tokens Output was $1.1176 per 1M tokens, now $2.574 per 1M tokens

Model: DeepSeek V4.1 Flash Input was $0.035 per 1M tokens, now $0.03483 per 1M tokens Output was $0.29 per 1M tokens, now $0.6 per 1M tokens

Model: DeepSeek Flash Latest Input was $0.035 per 1M tokens, now $0.03483 per 1M tokens Output was $0.29 per 1M tokens, now $0.6 per 1M tokens

Model: DeepSeek Pro Latest Input was $0.2442 per 1M tokens, now $0.24298 per 1M tokens Output was $0.7326 per 1M tokens, now $3.5 per 1M tokens

Model: DeepSeek V3 Input was $0.32 per 1M tokens, now $0.2574 per 1M tokens Output was $0.89 per 1M tokens, now $1.0287 per 1M tokens

Model: DeepSeek V4 Flash 0423 Input was $0.04704 per 1M tokens, now $0.0469 per 1M tokens Output was $0.09408 per 1M tokens, now $0.0938 per 1M tokens

Цены на API Gemini Developer | Gemini API | Google AI for Developers Перейти к основному контенту / English Deutsch Español – América Latina Français Indonesia Italiano Polski Português – Brasil Shqip Tiếng Việt Türkçe Русский עברית العربيّة فارسی हिंदी বাংলা ภาษาไทย 中文 – 简体 中文 – 繁體 日本語 한국어 Как по

Model: GLM Flash Latest Input was $0.04 per 1M tokens, now $0.045 per 1M tokens Output was $0.5 per 1M tokens, now $0.14 per 1M tokens

Model: MiniMax M2 Input was $0.3 per 1M tokens, now $0.255 per 1M tokens Output was $1.2 per 1M tokens, now $1.02 per 1M tokens

Model: MiniMax M2.7 Input was $0.3 per 1M tokens, now $0.21 per 1M tokens Output was $1.2 per 1M tokens, now $0.84 per 1M tokens

Model: DeepSeek V4 Pro 0813 Input was $0.264 per 1M tokens, now $0.24502 per 1M tokens Output was $0.792 per 1M tokens, now $3.5 per 1M tokens

Model: DeepSeek V4 Flash Vision Exp Input was $0.22 per 1M tokens, now $0.2156 per 1M tokens Output was $0.66 per 1M tokens, now $0.6468 per 1M tokens

Model: Kimi Latest Input was $1 per 1M tokens, now $1 per 1M tokens Output was $9 per 1M tokens, now $5.6 per 1M tokens

Model: GLM 5.3 Input was $0.3794 per 1M tokens, now $0.238 per 1M tokens Output was $1.1924 per 1M tokens, now $0.748 per 1M tokens

Model: GLM Latest Input was $0.3794 per 1M tokens, now $0.238 per 1M tokens Output was $1.1924 per 1M tokens, now $0.748 per 1M tokens

Model: DeepSeek Pro Latest Input was $0.24948 per 1M tokens, now $0.24684 per 1M tokens Output was $0.74844 per 1M tokens, now $0.74052 per 1M tokens

Model: DeepSeek Pro Latest Input was $0.25 per 1M tokens, now $0.24948 per 1M tokens Output was $3.5 per 1M tokens, now $0.74844 per 1M tokens

Model: GLM 5.3 Input was $1.4 per 1M tokens, now $0.3794 per 1M tokens Output was $4.4 per 1M tokens, now $1.1924 per 1M tokens

Model: Kimi Latest Input was $1.2 per 1M tokens, now $1.0301 per 1M tokens Output was $10.5346 per 1M tokens, now $9.043 per 1M tokens

Model: Nemotron 3.5 Lightning Input was $0.07 per 1M tokens, now $0.08 per 1M tokens Output was $0.2 per 1M tokens, now $0.2 per 1M tokens

Model: GLM Latest Input was $0.5614 per 1M tokens, now $0.4746 per 1M tokens Output was $1.7644 per 1M tokens, now $1.4916 per 1M tokens

Model: GLM 5.3 Flash Input was $0.045 per 1M tokens, now $0.04 per 1M tokens Output was $0.14 per 1M tokens, now $0.5 per 1M tokens

Model: GLM Flash Latest Input was $0.045 per 1M tokens, now $0.04 per 1M tokens Output was $0.14 per 1M tokens, now $0.5 per 1M tokens

Model: DeepSeek V4.1 Flash Input was $0.15 per 1M tokens, now $0.14 per 1M tokens Output was $0.6 per 1M tokens, now $0.42 per 1M tokens

Prezzi dell'API Gemini Developer | Gemini API | Google AI for Developers Passa ai contenuti principali / English Deutsch Español – América Latina Français Indonesia Italiano Polski Português – Brasil Shqip Tiếng Việt Türkçe Русский עברית العربيّة فارسی हिंदी বাংলা ภาษาไทย 中文 – 简体 中文 – 繁體 日本語

Model: DeepSeek V4 Pro 0423 Input was $0.570894 per 1M tokens, now $0.45936 per 1M tokens Output was $1.141788 per 1M tokens, now $0.91872 per 1M tokens

Model: Qwen3.6 27B Input was $0.32 per 1M tokens, now $0.32 per 1M tokens Output was $2.7 per 1M tokens, now $3.2 per 1M tokens

Model: DeepSeek V4 Flash 0731 Input was $0.03 per 1M tokens, now $0.022 per 1M tokens Output was $0.32 per 1M tokens, now $0.32 per 1M tokens

Model: DeepSeek V4 Flash Latest Input was $0.03 per 1M tokens, now $0.022 per 1M tokens Output was $0.32 per 1M tokens, now $0.32 per 1M tokens

Model: DeepSeek Flash Latest Input was $0.099 per 1M tokens, now $0.035 per 1M tokens Output was $0.6 per 1M tokens, now $0.29 per 1M tokens

Name: Jev Router Provider: TypeSafe Listed: 2026-09-25 Context window: 1000000 tokens Jev Router picks the best model and reasoning effort for each request, balancing quality, speed, and cost. It runs on [Jev](https://openrouter.ai/~typesafe/jev-latest), TypeSafe's first System One model, and adapts

Model: DeepSeek V4.1 Flash Input was $0.15 per 1M tokens, now $0.099 per 1M tokens Output was $0.6 per 1M tokens, now $0.6 per 1M tokens

Model: Kimi Latest Input was $0.8845 per 1M tokens, now $1.2 per 1M tokens Output was $10.5346 per 1M tokens, now $10.5346 per 1M tokens

Model: DeepSeek Flash Latest Input was $0.04 per 1M tokens, now $0.1 per 1M tokens Output was $0.49 per 1M tokens, now $0.5 per 1M tokens

Name: Perceptron Mk1.5 Provider: Perceptron Listed: 2026-09-25 Context window: 36864 tokens Input: $0.15 per 1M tokens Output: $1.5 per 1M tokens Perceptron Mk1.5 is Perceptron's perception model for physical agents. It accepts text, image, video, and audio input, and answers with text plus optional

Model: Gemma 4 26B A4B Input was $0.09 per 1M tokens, now $0.0675 per 1M tokens Output was $0.3 per 1M tokens, now $0.225 per 1M tokens

Model: DeepSeek V4 Flash 0423 Input was $0.049 per 1M tokens, now $0.04872 per 1M tokens Output was $0.098 per 1M tokens, now $0.09744 per 1M tokens

Model: Nemotron 3.5 Lightning Input was $0.08 per 1M tokens, now $0.07 per 1M tokens Output was $0.2 per 1M tokens, now $0.2 per 1M tokens

Model: DeepSeek V4 Pro 0813 Input was $0.462 per 1M tokens, now $0.34848 per 1M tokens Output was $1.386 per 1M tokens, now $1.04544 per 1M tokens

Model: Ling 3.0 Flash VL Input was $0.06 per 1M tokens, now $0.021 per 1M tokens Output was $0.18 per 1M tokens, now $0.0616 per 1M tokens

Model: DeepSeek V4 Pro 0423 Input was $0.79257 per 1M tokens, now $0.783 per 1M tokens Output was $1.58514 per 1M tokens, now $1.566 per 1M tokens

Model: Kimi K3 Input was $0.8845 per 1M tokens, now $3 per 1M tokens Output was $10.5346 per 1M tokens, now $15 per 1M tokens

Model: GLM 5.3 Flash Input was $0.045 per 1M tokens, now $0.045 per 1M tokens Output was $0.6 per 1M tokens, now $0.14 per 1M tokens

Model: DeepSeek Pro Latest Input was $0.26 per 1M tokens, now $0.09984 per 1M tokens Output was $3.5 per 1M tokens, now $2.88 per 1M tokens

Pricing | OpenAI API For the complete documentation index, see llms.txt . Markdown versions of documentation pages are available by appending .md to the page URL. ChatGPT Start searching API Dashboard Try ChatGPT Home API Codex Docs Guides, concepts, and product docs for Codex Use cases Example work

Gemini Developer API 가격 책정 | Gemini API | Google AI for Developers 기본 콘텐츠로 건너뛰기 / English Deutsch Español – América Latina Français Indonesia Italiano Polski Português – Brasil Shqip Tiếng Việt Türkçe Русский עברית العربيّة فارسی हिंदी বাংলা ภาษาไทย 中文 – 简体 中文 – 繁體 日本語 한국어 API 키 가져오기 설명서 커뮤니티 로그인

DeepSeek Pro Latest input pricing on OpenRouter has dropped from $0.3498 to $0.26 per 1M tokens. Output pricing has moved the other way, from $1.0494 to $3.5 per 1M tokens. The effective cost now depends heavily on how much the model writes rather than how much it reads.

Z.ai has raised the price of GLM 4.7 on OpenRouter. Input moves from $0.4 to $0.6 per 1M tokens, and output from $1.75 to $2.2 per 1M tokens. That is a 50% increase on input and roughly 26% on output.

Kimi K3 input pricing dropped from $3 to $0.8845 per 1M tokens, a cut of roughly 70%. Output pricing fell from $15 to $10.5346 per 1M tokens. The change is listed on OpenRouter's models endpoint.

DeepSeek V4.1 Flash pricing on OpenRouter has doubled. Input moves from $0.15 to $0.3 per 1M tokens, and output from $0.6 to $1.2 per 1M tokens.

DeepSeek Flash Latest output pricing on OpenRouter has dropped from $1 per 1M tokens to $0.49 per 1M tokens. Input pricing is unchanged at $0.04 per 1M tokens. The change affects output-heavy workloads most.

Google has published Gemini API pricing for Gemini 3.8 Flash with an introductory rate that ends on 31 December 2026. Paid tier standard input moves from USD 0.75 to USD 1.50 per million tokens on 1 January 2027, and output, including reasoning tokens, moves from USD 3.75 to USD 7.50. Batch, Flex and context caching rates double on the same schedule, and Priority input goes from USD 1.35 to USD 2.70.

Mistral Large 3 2512 pricing on OpenRouter has doubled across the board. Input moved from $0.25 to $0.5 per 1M tokens, and output moved from $0.75 to $1.5 per 1M tokens.

DeepSeek V4 Flash 0423 input pricing on OpenRouter dropped from $0.0686 to $0.049 per 1M tokens. Output dropped from $0.1372 to $0.098 per 1M tokens. Both sides fell by the same proportion, so the input to output ratio is unchanged.

Qwen3 VL 30B A3B Instruct pricing on OpenRouter has gone up. Input moves from $0.13 to $0.15 per 1M tokens, and output from $0.52 to $0.60 per 1M tokens.

Z.ai's GLM 5.1 pricing on OpenRouter moved down marginally. Input went from $0.966 to $0.9646 per 1M tokens, and output from $3.036 to $3.0316 per 1M tokens. The change is fractions of a cent per million tokens in both directions of the exchange.

DeepSeek V4 Pro 0423 input pricing on OpenRouter dropped from $0.844944 to $0.82128 per 1M tokens. Output pricing fell from $1.689888 to $1.64256 per 1M tokens. Both moves are small reductions of roughly 3%.

MoonshotAI has reduced pricing for Kimi Latest on OpenRouter. Input has dropped from $1.3 to $0.8845 per 1M tokens, and output from $11.4 to $10.5346 per 1M tokens. The input cut is the larger of the two in proportional terms.

Tencent's Hy3 input pricing moved from $0.0825 to $0.132 per 1M tokens, and output from $0.33 to $0.528 per 1M tokens, as listed on OpenRouter. Both rates rose by the same proportion, so the input-to-output ratio is unchanged.

DeepSeek Pro Latest input pricing has moved from $0.3894 to $0.3498 per 1M tokens, and output from $1.1682 to $1.0494 per 1M tokens, according to OpenRouter's model listing. Both figures are down roughly 10%.

A capture of the OpenAI API pricing page returned only site navigation and documentation index links, with no model rates, tiers or effective dates visible. The page itself remains the canonical location for OpenAI API pricing, and it notes that Markdown versions of docs pages are available by appending .md to the URL. No pricing change can be confirmed from this material.

DeepSeek V4 Pro 0423 input pricing on OpenRouter has moved from $0.871914 to $0.844944 per 1M tokens. Output pricing has moved from $1.743828 to $1.689888 per 1M tokens. Both are roughly a 3% reduction.

MoonshotAI has cut the listed price of Kimi Latest on OpenRouter. Input moves from $1.35 to $1.3 per 1M tokens, and output from $11.5 to $11.4 per 1M tokens. Both are small reductions rather than a repositioning of the model.

Z.ai has repriced GLM 5.3 Flash on OpenRouter. Input drops from $0.15 to $0.045 per 1M tokens, while output climbs from $0.5 to $0.6 per 1M tokens. The model is now much cheaper to feed and slightly more expensive to generate from.

Google's pricing page now lists Gemini 3.8 Flash, described as its most intelligent Flash model for long-horizon software engineering, autonomous agents and complex enterprise workflows. It is priced identically to Gemini 3.7 Flash on the paid tier, at $0.75 per 1M input tokens and $3.75 per 1M output tokens including thinking tokens through December 31, 2026, doubling to $1.50 and $7.50 from January 1, 2027. Batch and Flex tiers are half the standard rate, Priority is higher, and Grounding with Google Search and Google Maps give 5,000 free requests a month shared across Gemini 3.x models, then $14 per 1,000 requests.

Tencent's Hy3 has dropped in price on OpenRouter. Input moves from $0.132 to $0.0825 per 1M tokens, and output from $0.528 to $0.33 per 1M tokens. Both sides of the model fall by the same proportion.

DeepSeek V4.1 Flash pricing on OpenRouter has moved up on both sides. Input went from $0.14 to $0.15 per 1M tokens, a marginal change. Output went from $0.42 to $0.60 per 1M tokens, a rise of roughly 43%.

DeepSeek V4 Pro 0423 input pricing has moved from $0.888966 to $0.871914 per 1M tokens, with output dropping from $1.777932 to $1.743828 per 1M tokens. The change is listed on OpenRouter's models endpoint. Both figures are down by roughly two per cent.

OpenAI's API pricing page now sits inside the model catalog section of the platform docs, alongside model selection and choose-a-model guidance. The page also confirms markdown versions of any docs page are available by appending .md to the URL, with a full documentation index at llms.txt. No pricing figures are stated in the material we have.

DeepSeek V4 Flash 0423 input pricing has dropped from $0.08554 to $0.07476 per 1M tokens, and output from $0.17108 to $0.14952 per 1M tokens, according to OpenRouter's model listing. Both rates fall by roughly the same proportion, so the input to output ratio is unchanged.

MoonshotAI has adjusted pricing for Kimi Latest on OpenRouter. Input moved from $1.4 to $1.35 per 1M tokens, while output rose from $10.75 to $11.5 per 1M tokens.

DeepSeek V4 Pro 0423 input pricing on OpenRouter moved from $0.931074 to $0.919242 per 1M tokens. Output moved from $1.862148 to $1.838484 per 1M tokens. Both changes are a shade over one percent lower.

OpenAI has published a customer story on invideo building with GPT-6 Astra. invideo says the model plans edits with greater precision, improves colour correction and grading threefold, and produced 50 custom effects in a single day.

Harvey now uses GPT-6 Astra to generate legal documents. OpenAI says the model produces more structured, context-aware drafts, moving Harvey from assembling context to producing finished work lawyers can review rather than rebuild.

OpenAI has published the remarks its CEO Sam Altman gave to the United Nations Security Council. The remarks cover AI safety, human control and international cooperation. This is a policy communication, not a product or model change.

OpenAI is extending access to its Daybreak program to the Government of Ukraine. The stated purpose is supporting the cyber defence of civilian infrastructure. Access moves from a closed programme to including a national government partner.

OpenAI published a post marking two years of OpenAI Academy. The stated focus is bringing AI skills training to more communities. No product or model behaviour change is described.

Google Beam is expanding into five new countries, widening the set of markets where it is available. Google has also partnered with Industrious to extend the network of locations, alongside new partners and customers.

Name: Switchyard Provider: NVIDIA Listed: 2026-09-21 Context window: 1000000 tokens Switchyard is an open-source model router that switches between multiple models to optimize the cost of requests. By default it will use OpenRouter market data to select the most popular...

Google worked directly with designers Jane Wade and Sergio Hudson to custom-design Google Flow tools for their New York Fashion Week preparation. This moves Flow from a general-purpose tool to one shaped around a specific creative workflow through hands-on collaboration with practitioners.

Google is adding academic advisors, fellows and core internal researchers to its AI & Economy team. The change is to the team's composition, moving from an internal group to one supplemented by external academic expertise.

Google and the UN system have launched the UN System Data Commons, an open platform for global statistics. Data that was previously spread across UN sources is now searchable and explorable in one place.