DeepSeek Math 7B Instruct
DeepSeek Math 7B Instruct is a released coding and classification model with open-weight; evaluate it while provider pricing coverage matures.
Use it for
- Teams evaluating coding and classification
Do not use it for
- Cost-sensitive launches that need sourced token pricing
- Vision or document-understanding workloads
- Strict JSON or tool-calling flows
- Family
- DeepSeek Math
- Released
- 2024-02-05
- Parameters
- 7B
- Architecture
- Decoder Only
- Specialization
- math
- Openness
- Open weights
- License
- DeepSeek LicenseCommercial use: permitted
- Weights
- Available
- Code
- Unknown
- Training
- Fine-tuned
No tracked provider token pricing is available yet.
About
DeepSeek Math 7B Instruct is DeepSeek's DeepSeek Math model focused on mathematical reasoning. Weights are openly available for self-hosting and scores 49.2 on GPQA.
DeepSeek Math 7B Instruct is an open-weight model in the DeepSeek Math family. Headline tracked benchmarks include Google-Proof Q&A 49.2, HellaSwag 90.1, and HumanEval 78.9.
Top use-case fit: coding, agents, and build tasks
Coding
1 relevant benchmark in the decision map.
Classification
2 relevant benchmarks in the decision map.
Provider price ladder
No tracked provider token pricing is available for this model yet.
Capabilities
No model capability flags are currently sourced.
Benchmark peer barsfor Coding
Benchmark scores(4)
| Benchmark | Score | Version | Evaluation | Source |
|---|---|---|---|---|
| Google-Proof Q&A | 49.2 | diamondObserved 2026-03-06 | — | Source |
| HellaSwag | 90.1 | 10-shotObserved 2026-03-06 | — | Source |
| HumanEval | 78.9 | pass@1Observed 2026-03-06 | — | Source |
| Massive Multitask Language Understanding | 75.9 | 5-shotObserved 2026-03-06 | — | Source |
Migration checks
No linked migration route is available for this model yet.
Frequently asked questions
When was DeepSeek Math 7B Instruct released?
DeepSeek Math 7B Instruct was released on 2024-02-05.
What benchmarks has DeepSeek Math 7B Instruct been tested on?
DeepSeek Math 7B Instruct has been evaluated on 4 benchmarks, including Google-Proof Q&A, HellaSwag, HumanEval, Massive Multitask Language Understanding.