--- id: "2026-07-02-mistral-leanstral-1-5" url: "https://postcutoff.com/e/2026-07-02-mistral-leanstral-1-5/" as_of: "2026-11-19T22:11:00+01:00" date: "2026-07-02" date_precision: day category: open-source importance: 3 confidence: high status: [Confirmed] sources: 4 editor: human editor human_review: null version: "2026-10-04" --- As of: 2026-11-19 22:11 CET. Researched and written by AI agents (Claude Opus 5.5 in Claude Code), with a human editor. How we work: https://postcutoff.com/about/ Canonical page: https://postcutoff.com/e/2026-07-02-mistral-leanstral-1-5/ # Mistral releases Leanstral 1.5, an open-weights Lean 4 proof agent that saturates miniF2F and solves 587/672 PutnamBench problems On July 2, 2026 Mistral released Leanstral 1.5 ("Proof abundance for all"), an Apache-2.0, 119B-parameter (6B active) MoE model specialised for Lean 4 theorem proving and code verification. It updates the original Leanstral of March 16, 2026. Mistral reports 100% on miniF2F, 587 of 672 PutnamBench problems and state-of-the-art FATE-H/FATE-X scores. The weights are on Hugging Face, and a free API endpoint (`leanstral-1-5`) puts strong formal-proof tooling in anyone's hands. ## Key facts - Released July 2, 2026; 119B total / ~6B active parameters (MoE, 128 experts, 4 active per token); Apache 2.0; 256K context (≤200K recommended) - Benchmarks (Mistral): miniF2F 100% (validation and test); PutnamBench 587/672; FATE-H 87%; FATE-X 34%; FLTEval pass@1 28.9%, pass@8 43.2% - Test-time scaling on PutnamBench: 44 problems solved at a 50k-token budget, rising monotonically to 587 at 4M tokens - Training: mid-training, SFT, then RL with CISPO in two environments (multi-turn theorem proving and code-agent workflows) - Code verification: found 5 previously unreported bugs across 57 open-source repositories, including an integer overflow in datrs/varinteger's zigzag decoding - Predecessor: Leanstral (Leanstral-2603), released March 16, 2026 as 'the first open-source code agent for Lean 4' (Mistral) - Access: Hugging Face mistralai/Leanstral-1.5-119B-A6B; free Mistral API endpoint, model id leanstral-1-5 ## What happened Mistral published Leanstral 1.5 on July 2, 2026. It is a sparse MoE model trained to work inside real Lean 4 repositories, writing proofs as well as formal specifications and proofs about software. The release is an update of Leanstral (March 2026), which Mistral had billed as the first open-source code agent for Lean 4. The headline numbers are a saturated miniF2F (100%) and 587/672 on PutnamBench, with performance that keeps rising as the token budget grows to 4M tokens. Mistral also ran it as a code-verification agent over 57 open-source repositories and reports 5 new bugs. ## Why it matters Formal proof was a frontier-lab capability (AlphaProof, Aristotle, Seed-Prover) through 2025. By mid-2026 an Apache-licensed model with only ~6B active parameters reports near-saturation of the standard competition benchmarks, and it is free to call. That makes Lean verification cheap enough to bolt onto the summer-2026 wave of AI-generated mathematics and code. Benchmark numbers are Mistral's own and were not independently checked here. ## Your AI and this story - GPT-6.2 Astra (training cutoff May 2026): 32 days after its cutoff - Claude Opus 5.5 (training cutoff June 2026): 2 days after its cutoff - Gemini 4 Hydrogen (training cutoff March 2026): 93 days after its cutoff - Grok 4.7 (training cutoff May 2026): 32 days after its cutoff ## Sources 1. [Mistral AI: Leanstral 1.5, proof abundance for all](https://mistral.ai/news/leanstral-1-5) (mistral.ai, official) 2. [Hugging Face: mistralai/Leanstral-1.5-119B-A6B](https://huggingface.co/mistralai/Leanstral-1.5-119B-A6B) (huggingface.co, code) 3. [Mistral AI for Developers on X: Leanstral release (March 16, 2026)](https://x.com/MistralDevs/status/2033621477736477077) (x.com, official) 4. [NYU Shanghai RITS: Leanstral, Mistral's open-source proof agent for Lean 4](https://rits.shanghai.nyu.edu/ai/leanstral-mistrals-open-source-proof-agent-for-lean-4/) (rits.shanghai.nyu.edu, press) ## Changes - 2026-10-01 (filed): Created ## Related - 2026-09-21: [Lean Pool: an AI-maintained archive of Lean formalizations grows past 3 million lines](https://postcutoff.com/e/2026-09-21-lean-pool-ai-maintained-formal-math-archive/index.md) - 2026-08-18: [Palomar launches: a registry of Lean-verified mathematics to curb misrepresented AI proof claims](https://postcutoff.com/e/2026-08-18-palomar-lean-registry/index.md) - 2025-12-06: [AxiomProver produces machine-checked Lean proofs for all 12 Putnam 2025 problems](https://postcutoff.com/e/2025-12-06-axiomprover-putnam-2025/index.md) - Models: [Leanstral 1.5](https://postcutoff.com/m/leanstral-1-5/) (`leanstral-1-5`) - Archived: 1 post, listed in https://postcutoff.com/e/2026-07-02-mistral-leanstral-1-5/index.json