Requesty
Fireworks AI

minimax-m3

MiniMax-M3 is a native multimodal model with 512K context running ~428B parameters and ~23B activated parameters. It brings native multimodality, enabling deeper semantic fusion across text, image, and video. M3 also introduces MiniMax Sparse Attention (MSA) to improve long context efficiency, achieving frontier-level performance across long-horizon agentic benchmarks, excelling in both coding and cowork.

VisionTool callingCaching
Try minimax-m3CompareDocs

Specifications

Context window512K tokens
Max output512K tokens
API typechat
AddedJun 13, 2026
Model IDfireworks/minimax-m3
Data retentionNo
Used for trainingNo
Provider location🇺🇸 US
Privacy policyFireworks Privacy Policy→

Benchmarks

Released 2026-06-01
Coding Indexcoding
43.4%

Artificial Analysis Coding Index — a composite of coding evaluations including LiveCodeBench, SciCode and Terminal-Bench.

GPQA Diamondreasoning
92.9%

Graduate-level physics, chemistry & biology questions designed to resist Googling.

Intelligence Indexreasoning
54.7%

Artificial Analysis Intelligence Index — a composite of multiple evaluations measuring overall model capability.

Scores are sourced from official model cards, Artificial Analysis, and public leaderboards. Benchmarks measure specific skills and do not capture every aspect of model quality. Always test on your own workload.

Pricing

Input / 1M
$0.30
Output / 1M
$1.20
Cache write
N/A
Cache read / 1M
$0.06
Estimated cost
100K input + 10K output$0.0420
1M input + 100K output$0.42
10M input + 1M output$4.20

Requesty charges exactly what the upstream provider charges, no markup, no per-request fees. Prompt caching and smart routing can reduce effective cost by 30-80%.

Quickstart

Drop-in compatible with the OpenAI SDK. Change the base URL, swap in your Requesty API key, and set the model to fireworks/minimax-m3.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
from openai import OpenAI

client = OpenAI(
    api_key="YOUR_REQUESTY_API_KEY",
    base_url="https://router.requesty.ai/v1",
)

response = client.chat.completions.create(
    model="fireworks/minimax-m3",
    messages=[
        {"role": "user", "content": "Explain quantum computing in one paragraph."},
    ],
)

print(response.choices[0].message.content)

Other Fireworks AI models

qwen3.7-plus
262K ctx$0.40/1M
Coding 47
nemotron-3-ultra-nvfp4
262K ctx$0.60/1M
kimi-k2.7-code
262K ctx$0.95/1M
gpt-oss-120b
131K ctx$0.15/1M
Coding 29
gpt-oss-20b
131K ctx$0.07/1M
Coding 19
deepseek-v4-flash
1M ctx$0.14/1M
Coding 39

Frequently asked questions

How much does minimax-m3 cost?
minimax-m3 is priced at $0.30 per million input tokens and $1.20 per million output tokens when accessed via Requesty. Prompt caching is supported, which can cut effective input cost by up to 90% on repeated context. Requesty charges exactly what the upstream provider charges, we don't add markup.
What is the context window of minimax-m3?
minimax-m3 has a context window of 512K tokens, with a maximum output of 512K tokens per response. That's roughly 683 words of input you can fit in a single prompt.
How does minimax-m3 perform on benchmarks?
minimax-m3 scores 92.9% on GPQA Diamond, 88.9% on τ²-Bench, 54.7% on Intelligence Index. See the full benchmark chart above for results across MMLU Pro, GPQA Diamond, SWE-Bench Verified, HumanEval, MATH, AIME, MMMU, and LiveBench.
What can minimax-m3 do?
minimax-m3 supports vision input, tool calling, prompt caching. You can call it through any OpenAI-compatible client by pointing base_url to Requesty.
How do I use minimax-m3 with the OpenAI SDK?
Install the OpenAI SDK, set base_url to "https://router.requesty.ai/v1", set your API key to your Requesty key, and set the model to "fireworks/minimax-m3". The Quickstart above shows Python, JavaScript and cURL snippets.
Can I run minimax-m3 through Requesty?
Yes. minimax-m3 runs through Requesty's OpenAI-compatible API, served from Fireworks AI. You do not host the model yourself: point base_url at Requesty, set the model to "fireworks/minimax-m3", and requests are routed to the upstream provider with automatic failover. The same key gives you 400+ other models too.

Access minimax-m3 through Requesty

One API key, 400+ models, OpenAI-compatible. No markup on provider prices, automatic failover, and smart caching built-in.

Get started freeAll Fireworks AI models