See how ChatGPT, Perplexity and Google AI Overviews describe you today
Free AI Visibility AuditMiniMax M1 is an open-weight reasoning model from MiniMax, built for long-context work and released in June 2025. It is a text-only model that takes a very large input window and is priced at the cheap end of the reasoning tier.
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
$0.40
field median $0.30
Output / 1M tokens
$2.20
field median $1.25
Context
1,000K
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.
MiniMax M1 is a reasoning model, meaning it works through a problem in steps before it answers, and its weights are published rather than held behind an API only. It was built for jobs where the input is long and the reasoning has to hold together across all of it, such as reading a whole codebase, a document set or a long transcript. It handles text alone, so it is not the model for anything involving images, audio or video.
Where it sits in its line, and when a sibling is the better pick.
MiniMax M1 first appeared in June 2025 as MiniMax's reasoning-focused open-weight release, sitting above the company's general chat models on anything that needs deliberate multi-step work. Pick it when the task is long-input analysis and you want the option of running the weights yourself. Pick something else when you need images or audio in the prompt, or when a very long answer is the point, since its output ceiling is modest next to its input capacity.
The API, the apps it powers, and what its limits let you do.
MiniMax M1 is reached through MiniMax's API and through the inference hosts that carry open-weight models, and because the weights are published it can also be run on your own hardware. Its context window stretches to a million tokens, so a full repository, a year of support tickets or a long contract set can go in as a single prompt without a retrieval layer in front of it. The output cap is much smaller than the input, which suits analysis, extraction and code edits rather than long-form generation.
What changes because this exists, or why it does not.
The combination here is the point: a reasoning model with a million-token window, open weights and a price close to a standard chat model. Work that previously meant chunking documents, building a retrieval pipeline and stitching partial answers together can be done in one call, and teams with data they cannot send to a third party can host it themselves rather than dropping the capability.
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 | $0.40 | per 1M tokens | 2026-09-24 | Check |
| Output | $2.20 | per 1M tokens | 2026-09-24 | Check |
MiniMax M1 is inexpensive to run, sitting near the bottom of the reasoning tier and well below the frontier closed models it competes with on long-context tasks. Output costs several times more than input, which matters less here than usual because the model is built to read a great deal and answer briefly, and the open weights mean self-hosting is a route to cutting the per-call cost further.
| Workload | Input / month | Output / month | Cost |
|---|---|---|---|
| A small product team | 20M tokens | 5M tokens | $19.00 |
| A busy support assistant | 200M tokens | 40M tokens | $168.00 |
| A document pipeline | 1000M tokens | 100M tokens | $620.00 |
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 | 1,000,000 tokens |
|---|---|
| Maximum output | 40,000 tokens |
| Modalities | text |
| Released | 17/06/2025 |
| Status | Current |
| Catalogue identifier | minimax/minimax-m1 |
08 / lineage
What this model replaced, what replaced it, and what else its provider has in the field.
Also from MiniMax
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
Open-weight models like MiniMax M1 mean your brand can be summarised by a system nobody is publicly tracking, running inside somebody else's product or on a customer's own hardware. You cannot monitor those surfaces the way you monitor a named assistant, so the work shifts to what the model reads: clear, well-structured pages that survive being ingested in bulk alongside a million tokens of competitor material. Long-context reasoning also rewards depth, since a model taking in your entire documentation set will reward the brand whose material actually answers the question end to end.
Most buyers meet MiniMax M1 indirectly, through a product that has chosen it as a cheap long-context reasoning backend rather than through a consumer app with its name on it. Developers meet it through the MiniMax API or an open-weight inference host, and increasingly through self-hosted deployments inside companies that will not send documents out.
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.