As of: 2026-11-13 13:49 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/m/deepseekmath-v2/

# DeepSeekMath-V2

DeepSeek, DeepSeekMath family. Status: current. Type: reasoning model. Released: 2025-11-27. Training cutoff: not published.

## How to call it

| Provider | Model id | Endpoint | Docs |
|---|---|---|---|
| Hugging Face |  |  | https://huggingface.co/deepseek-ai/DeepSeek-Math-V2 |

## Pricing

- Price not published.
- Last checked: 2026-09-29

## Specs

- Input: text
- Output: text
- Context window: not published
- Max output: not published
- Open weights: yes
- Licence: apache-2.0

## What it has not seen

- No published training cutoff. Counted from its release (2025-11-27): at least 1,970 AI events since (450 major, 85 historic), as of 2026-11-13.
- Everything since October 2025: https://postcutoff.com/since/2025-10/

## What stands out

- Self-verifiable proof generation (first): Generator trained against an LLM proof verifier and meta-verifier; reached IMO 2025 / CMO 2024 gold level and 118/120 on Putnam 2024 with scaled test-time compute. Source: https://arxiv.org/abs/2511.22570

## Notes

685B open-weights (Apache 2.0) math prover built on DeepSeek-V3.2-Exp-Base; inference uses the DeepSeek-V3.2-Exp code. No first-party API endpoint verified. 'first' = first open-weights model at IMO-gold level (per the paper's claims).

Download the weights from Hugging Face and serve them with the DeepSeek-V3.2-Exp inference stack. Source: https://huggingface.co/deepseek-ai/DeepSeek-Math-V2 · paper https://arxiv.org/abs/2511.22570

Model page: https://postcutoff.com/m/deepseekmath-v2/