See how ChatGPT, Perplexity and Google AI Overviews describe you today
Free AI Visibility AuditKimi K2 0905 is a text-only, open-weight large language model from MoonshotAI, released as the September update to the earlier Kimi K2 0711. It is a Mixture-of-Experts model built at trillion-parameter scale, with a long context window and unusually cheap token pricing for its size.
01 / decision
Each figure sits next to the middle of the field, so it can be read as dear or cheap, wide or narrow, rather than floating on its own. Compared against the 126 models we track, not against an absolute standard.
Capability
Not measured
No published benchmark scores yet
Input / 1M tokens
$0.60
field median $0.30
Output / 1M tokens
$2.50
field median $1.25
Context
262K
field median 524K
02 / overview
Four questions, answered separately, because somebody arrives at one of them rather than at the top of the page.
The job it was built for, and the job it is not for.
Kimi K2 0905 is a general-purpose text model built on a Mixture-of-Experts design, where only a fraction of its total parameters are active on any given request. The job it is built for is high-volume text work at long context: reading large documents, working across a big codebase or corpus, and producing long outputs in a single pass. It handles text only, so images, audio and video are out of scope.
Where it sits in its line, and when a sibling is the better pick.
Kimi K2 0905 first appeared in early September 2025 as the successor to Kimi K2 0711 in MoonshotAI's K2 line. It is the right pick over the July release when you want the current version of the same model family without changing your integration. It is the wrong pick when you need image or audio input, or when you are tied to a closed frontier model for reasons of procurement or vendor support.
The API, the apps it powers, and what its limits let you do.
Kimi K2 0905 is reached through an API, either MoonshotAI's own or one of the aggregators that route to it, and the weights are published so it can also be self-hosted. The context window is large enough to hold a long document set, a full support history or a substantial repository in one request, and the generous output ceiling means long reports, migrations and refactors do not have to be chunked. Everything in and out is text.
What changes because this exists, or why it does not.
Kimi K2 0905 makes long-context, high-volume text pipelines affordable to run continuously rather than as an occasional batch. Work that was awkward before, feeding an entire corpus in rather than building a retrieval layer around it, or generating a very long document in one call instead of stitching sections, becomes a simple engineering decision rather than a cost decision. As an open-weight release it also gives teams a self-hosting path if they need it.
03 / evidence
Split by category, so a strong number never hides a thin evidence base. Verified means we read it on the benchmark's own published results; a provider's claim about its own model is shown and labelled rather than dropped.
No published benchmark scores for this model yet.
We publish a score only where we can link the result to where it was published. Until a benchmark result for this model exists in a source we read, this section stays empty rather than being filled with a provider's marketing figure.
How we decide what counts as evidence04 / ledger
Every published row, grouped by category, each compared with the best published score on the same benchmark. 'Is 64% good' is a question nobody can answer; '26 points behind the leader' is one anybody can.
Nothing in the ledger yet.
Each row here carries a score, the benchmark it came from, what the leading model scored on the same test, and a link to the published result. Rows appear as results are published and read.
How we decide what counts as evidence05 / capability
Where this model is strong, and against how many peers. Ranks are against models with evidence in that category, not against everything we track: ranking against models nobody tested would rank who published, not who is better.
No category scores to shape yet.
A category score is the weighted mean of the benchmarks published for it. With no published rows there is nothing to average, and an empty chart drawn at zero would say something false.
How we decide what counts as evidence06 / cost
List API rates as last read from the provider, with the source on every row, plus every change we have recorded since we started tracking it.
| Charge | Price | Unit | Read on | Source |
|---|---|---|---|---|
| Input | $0.60 | per 1M tokens | 2026-09-24 | Check |
| Output | $2.50 | per 1M tokens | 2026-09-24 | Check |
Kimi K2 0905 is priced at the cheap end of the market, with output costing a few times more than input, the usual pattern. For a model of this scale and context length it undercuts the closed frontier models by a wide margin, which is the main reason teams put it into high-volume pipelines.
| Workload | Input / month | Output / month | Cost |
|---|---|---|---|
| A small product team | 20M tokens | 5M tokens | $24.50 |
| A busy support assistant | 200M tokens | 40M tokens | $220.00 |
| A document pipeline | 1000M tokens | 100M tokens | $850.00 |
List API rates, no caching and no batch discount, which both providers offer and which change the answer a great deal. Treat these as the ceiling, not the bill.
07 / specs
As listed by the provider's own catalogue and re-read every few hours. Anything absent is absent there too.
| Context window | 262,144 tokens |
|---|---|
| Maximum output | 98,304 tokens |
| Modalities | text |
| Released | 04/09/2025 |
| Status | Current |
| Catalogue identifier | moonshotai/kimi-k2-0905 |
08 / lineage
What this model replaced, what replaced it, and what else its provider has in the field.
Also from Moonshot AI
09 / line
Every model in this family in release order, so a page from eight months ago says in one glance that two newer ones exist.
Nothing else in this line yet.
A line is worked out from the naming and the release dates across every model page we hold. It fills in as the provider ships successors, or as we pick up the models that came before this one.
How we decide what counts as evidence10 / notes
What this model changes for a brand trying to be cited in AI answers, and every change we have logged since it launched.
What it changes for you
Models like Kimi K2 0905 lower the cost of reading a lot of text, which means the systems summarising and comparing vendors can afford to ingest far more of your site and documentation before answering. Thin pages and marketing copy get diluted at that volume, detailed product documentation, pricing explanations and comparison pages get picked up. If your material is only skimmable, it will be skimmed past in favour of a competitor who wrote the full version.
Most buyers meet Kimi K2 0905 through the API rather than a consumer app, usually inside a product someone else has built on it, or through a model-routing platform where it sits as one option among many. Developers encounter it directly when picking a cheap long-context model for a batch or agent workload. It is not the default engine behind a mainstream Western AI search surface, so it shapes what buyers see indirectly, through the tools built on top of it.
Nothing published here yet. Changes appear within hours of a provider announcing them.
11 / questions
The things people ask about this model, answered from what is on this page rather than from anywhere else.
Models change what gets cited. We measure whether AI answers name your brand or your competitors across every assistant, and show you what to change.