Mistral AI
Mistral AI's frontier-class open-weight multimodal model optimized for agentic and coding use cases, with a 256K context window.
Running this yourself: likely needs a high-memory cloud gpu.
OpenRouter
Price record: 2026-11-17. Source: openrouter.
No current subscription pricing is tracked for this model.
Confirm this specific model, usage limits, and billing terms with the provider. A subscription does not automatically include API credits.
No login needed to compare. Prices are in USD; provider charges are separate from AI Market Cap plans. Context length, caching, tools, taxes, and regional terms can change the final cost. Open weights do not mean free hosting.
44.6
Quality Score
1200
Arena ELO
128B
Parameters
262K
Context
This measures the amount of verifiable public evidence we have, not how capable the model is. A missing field means it has not been verified yet, not that its value is zero.
19 of 22 public signals
Sign in to join the discussion
0
Downloads
0
Likes
Apr 2026
Released
5/5 signals
4/4 signals
5/5 signals
1/4 signals
4/4 signals
Parameters
128B
Training compute
Not reported
Dataset scale
Not reported
Base model
Not reported
Source-reported access: Open weights (restricted use) · Confident confidence
Gaps we are still tracking
Launches
2
Benchmarks
20
Open Source
1
Research
2
General
5
Recent launch, pricing, benchmark, and API signals linked to this model or its provider.
Quality: 4.6/100 | Price: $3/M tokens | Output: 168.38 tok/s | HumanEval: 0.402%
Today, we’re introducing the Responses API in public preview, made for building agents on Mistral. Each response returns the model’s answer, its reasoning and its tool calls as separate items, so your app can handle every step more easily than with Chat Completions. It works https://t.co/5zgjZgqugd
View sourceQuality: 4.6/100 | Price: $3/M tokens | Output: 168.38 tok/s | HumanEval: 0.402%
View sourcePrice: $3/M tokens | Output: 168.642 tok/s | HumanEval: 0.402%
View sourcePrice: $3/M tokens | Output: 165.218 tok/s | HumanEval: 0.402%
View sourceWhat does real-time transcription look like in a live setting? 🎙 We used Voxtral Transcribe 3 at AI Engineer Paris to capture every main-stage talk in real time. We boosted accuracy with context biasing, feeding the model speaker names and technical terms from the conference https://t.co/L2mxmjrlav
What does real-time transcription look like in a live setting? We used Voxtral Transcribe 3 at AI Engineer Paris to capture every main-stage talk in real time. We boosted accuracy with context biasing, feeding the model speaker names and technical terms from the conference https://t.co/UKbQD6DJPk
What does real-time transcription look like in a live setting? 🎙We used Voxtral 3 Transcribe at AI Engineer Paris to capture every main-stage talk in real time. We boosted accuracy with context biasing, feeding the model speaker names and technical terms from the conference https://t.co/Fgf1Py4Z6I

Today, we release the weights of a new and improved Le Chonk, advancing the open frontier for agentic coding and cybersecurity. 💻 Built to build: Le Chonk ranks #1 on @cognition’s FrontierCode among open models, and sets a new benchmark for Pareto efficiency on https://t.co/R9wQ6WTPnl

Today, we’re introducing the Responses API in public preview, made for building agents on Mistral. Each response returns the model’s answer, its reasoning and its tool calls as separate items, so your app can handle every step more easily than with Chat Completions. It works https://t.co/5zgjZgqugd

Today we’re announcing a new collaboration with @nvidia to bring AI to industrial engineering: - faster simulation - open benchmarks and supporting datasets for physics AI - evaluations for agentic AI in CAE and EDA More here: https://t.co/HXFVBqJSP1 https://t.co/OTb24yjOUQ

What can large-scale AI models bring to the growing complexity of particle physics? @CERN and Mistral join forces to explore this question. The ambition: a single Large Physics Model, trained on billions of simulated collisions. https://t.co/nrdCfcBI3k
Price: $3/M tokens | Output: 168.642 tok/s | HumanEval: 0.402%
Price: $3/M tokens | Output: 165.218 tok/s | HumanEval: 0.402%
Price: $3/M tokens | Output: 160.839 tok/s | HumanEval: 0.402%
Price: $3/M tokens | Output: 159.714 tok/s | HumanEval: 0.402%
Price: $3/M tokens | Output: 151.515 tok/s | HumanEval: 0.402%
Price: $3/M tokens | Output: 156.516 tok/s | HumanEval: 0.402%
Price: $3/M tokens | Output: 161.743 tok/s | HumanEval: 0.402%
Quality: 14.2/100 | Price: $3/M tokens | Output: 164.95 tok/s | HumanEval: 0.402%
Quality: 14.2/100 | Price: $3/M tokens | Output: 170.747 tok/s | HumanEval: 0.402%
Quality: 14.2/100 | Price: $3/M tokens | Output: 158.085 tok/s | HumanEval: 0.402%
Quality: 14.2/100 | Price: $3/M tokens | Output: 159.278 tok/s | HumanEval: 0.402%
Quality: 14.2/100 | Price: $3/M tokens | Output: 156.771 tok/s | HumanEval: 0.402%
Quality: 14.2/100 | Price: $3/M tokens | Output: 154.197 tok/s | HumanEval: 0.402%
Quality: 14.2/100 | Price: $3/M tokens | Output: 151.463 tok/s | HumanEval: 0.402%
Quality: 14.2/100 | Price: $3/M tokens | Output: 155.479 tok/s | HumanEval: 0.402%
Quality: 14.2/100 | Price: $3/M tokens | Output: 154.537 tok/s | HumanEval: 0.402%
Quality: 14.2/100 | Price: $3/M tokens | Output: 155.546 tok/s | HumanEval: 0.402%
Quality: 14.2/100 | Price: $3/M tokens | Output: 166.611 tok/s | HumanEval: 0.402%
Large Language Model inference over long contexts is increasingly bottlenecked by the cost of attending to a growing Key-Value (KV) cache. Sparse attention reduces this cost by accessing only a fraction of the cache, but existing methods face a trade-off between deterministic retrieval, which captures high-scoring tokens but discards residual attention mass, and sampling, which improves coverage at the cost of higher variance. Hybrid methods combine top-k retrieval with uniform sampling, yet still exhibit a substantial gap from dense attention at aggressive sparsity. We introduce Stratified Sparse Attention (SSA), a unified framework combining deterministic retrieval with score-guided sampling. SSA deterministically attends to high-scoring tokens while sampling the residual using proxy attention scores and correcting for their inclusion probabilities, thereby recovering attention mass beyond the retrieval set. SSA encompasses top-k, top-p, and hybrid top-k+sampling as special cases. Across RULER, LongBench, LOFT, and AIME, and five models ranging from 3B to 27B parameters, SSA consistently improves over the baselines at matched sparsity. On RULER-HARD with Ministral-3-8B, SSA improves accuracy over the strongest practical baseline from 81.8\% to 84.2\% at 32K and from 63.8\% to 69.3\% at 128K under 50times attention sparsity, recovering 94\% and 85\% of dense-attention performance, respectively. With optimized CUDA kernels, SSA achieves up to 2.96times decode speedup over FlashInfer at 1M context, demonstrating that its accuracy gains translate into practical long-context inference acceleration. We open-source our code at https://github.com/skylight-org/SSA-Paper-Code.
Mistral Medium 3.5 is now available through local Ollama runtime. 256K context window listed. Mistral Medium 3.5 is the first flagship model of Mistral AI that merged instruction-following, reasoning, and coding in a single set of 128B weights.