MODEL PROOF
MiniMax M3
BETTER VALUE OPTION
A stored alternative has an equal-or-higher measured score and an equal-or-lower price.
Serving precision differs between offers.
GLM 5.3 Flash · Δ score 37.5 points · 51% lower measured price · $0.238 / 1M tokens
DeepSeek V4.1 Flash is a further stored option with these losses: no video input
- Independent LMArena score
- 1,432.1 ± 3.92 · 58,757 votes
- Context window
- 1.05M
- Maximum output
- 512K
- Input modalities
- text, image, video
- Output modalities
- text
- Published input price
- $0.280 / 1M tokens
- Published output price
- $1.10 / 1M tokens
- Pricing kind
- fixed
Inspect complete billing conditions and endpoint terms below.
Price from DeepInfra: the cheapest offer that is serving, at standard delivery and at the maker’s precision or better, at its standard rate.
Cheaper right now: $0.413/M at CoreWeave — 4-bit It does not pass the like-for-like test, so it never sets a rank.
Every price dimension · USD per million tokens
| Input | $0.280/M |
|---|---|
| Output | $1.10/M |
| Cached input | $0.056/M |
| Cache write | Unknown/M |
| Cache write 1h | Unknown/M |
| Reasoning | Unknown/M |
Who sells it
The same weights, different shops. Cheapest is not like-for-like when serving precision differs.
Headline cheapest is a lower precision. Like-for-like at the best declared precision is $0.280/M from DeepInfra.
13 provider offers · rates, limits and conditions
Rates are USD per million tokens. Endpoint terms can differ even when the quoted price and precision match. A listed parameter is a provider declaration, not a task-success test.
| Seller | Input | Output | Precision | Uptime (1d) |
|---|---|---|---|---|
| CoreWeavecoreweave/fp4 Endpoint terms
| $0.230 | $0.960 | fp4 | 98.47% |
| GMICloudgmicloud/fp8 Endpoint terms
| $0.240 | $0.960 | fp8 | 99.22% |
| DeepInfradeepinfra/fp8 Endpoint terms
| $0.280 | $1.10 | fp8 | 99.86% |
| StreamLakestreamlake/fp8 Endpoint terms
| $0.300 | $1.20 | fp8 | 98.69% |
| Venicevenice/fp8 Endpoint terms
| $0.300 | $1.20 | fp8 | 97.21% |
| Parasailparasail/fp8 Endpoint terms
| $0.300 | $1.20 | fp8 | 99.86% |
| AtlasCloudatlas-cloud/fp8 Endpoint terms
| $0.300 | $1.20 | fp8 | 99.62% |
| Novitanovita/fp8 Endpoint terms
| $0.300 | $1.20 | fp8 | 99.44% |
| Minimaxminimax/fp8 Endpoint terms
| $0.300 | $1.20 | fp8 | 99.67% |
| Togethertogether Endpoint terms
| $0.300 | $1.20 | undeclared | 99.26% |
| Maramara Endpoint terms
| $0.600 | $2.40 | undeclared | 93.42% |
| SambaNovasambanova Endpoint terms
| $0.600 | $2.40 | undeclared | 96.17% |
| ModelRunmodelrun/fp4 Endpoint terms
| $0.750 | $3.00 | fp4 | 97.84% |
Independent scores
| LMArena Elo | 1432.1default |
|---|
LMArena Elo under CC BY 4.0. Absence of another board is not a score of zero.
Best rankings by task
- Htmlslides #12 1180
- Python pptxslides #14 1204
- Full-stack apps #18 1196
- Agenticgamedev #19 1158
- Web apps #21 1192
- Mobile apps #23 1167
- Native Android apps #23 1171
- Websites #31 1264
Per-task Elo and rank from Design Arena, via OpenRouter’s model feed. The rank is the feed’s, and it counts models OpenRouter does not list.
Capability
| Context window | 1M |
|---|---|
| Max output | 512K |
| Input modes | text, image, video |
| Tool use | yes |
| Reasoning | optional |
| Open weights | yes |
Provenance
| Price source | openrouter.ai |
|---|---|
| Fetched | 2026-10-06 |
| Quality data | verified |
| Cross-checked | vendor page |
People search MiniMax. This page is MiniMax M3, not M2.7 and not a MiniMax subscription. This catalogue does not publish a parameter count. A missing count is not a guess. Artificial Analysis figures are not published here. Quality on this page is LMArena Elo.
MiniMax-M3 is a multimodal foundation model from MiniMax. It supports text, image, and video inputs with text output, a 1M-token context window, and is suited for long-horizon agentic work, coding,...
Questions this page answers
What does MiniMax M3 cost?
$0.280/M in, $1.10/M out in this catalogue, as of the fetch date on this page. That is the API row, not a subscription.
Does MiniMax M3 have an independent quality score?
MiniMax M3 has an LMArena score in this catalogue.
What beats MiniMax M3?
GLM 5.3 Flash has an equal-or-higher measured score and an equal-or-lower price.
Does this page use Artificial Analysis scores for MiniMax M3?
No. Artificial Analysis figures are not published here. Quality on this page is LMArena Elo where a score exists; otherwise the row is unrated.
Is the cheapest MiniMax M3 endpoint the same product?
Headline cheapest is a lower precision. Like-for-like at the best declared precision is $0.280/M from DeepInfra.
How many parameters does MiniMax M3 have?
This catalogue does not publish a parameter count for MiniMax M3. A missing count is not a guess.
MODEL MONUMENT
Save or share this proof
Badge
The verdict as one SVG, for a README or a docs page. It states this model's dominance status at the balanced workload on the LMArena lens, with the date it was computed, and it is rebuilt with the catalogue — so it changes when the verdict changes, including to one you would rather it did not.
[](https://undominated.ai/models/minimax__minimax-m3/)<a href="https://undominated.ai/models/minimax__minimax-m3/"><img src="https://undominated.ai/badge/minimax__minimax-m3.svg" alt="MiniMax M3 — Undominated.ai dominance verdict" height="36"></a>Direct file: https://undominated.ai/badge/minimax__minimax-m3.svg — a static SVG, written by scripts/build-badges.mjs on every build, so the copy you embed is never
older than the last deploy.
