Model
Model explorer
DeepSeekMath 7B
OPENDeepSeek · DeepSeekMath family · released Feb 5, 2024
Debuted GRPO reinforcement learning (later reused for R1); DeepSeekMath-RL 7B hit 51.7% on MATH via CoT, near GPT-4/Gemini-Ultra level at just 7B.
ReasoningCodingVisionFunction callingTool useAgentic
712.5
Elo · rank #277
Parameters
7B
Active params
7B (dense)
Context
4K tokens
Architecture
Dense transformer, initialized from DeepSeek-Coder-Base-v1.5 7B
License
DeepSeek Model License
Languages
—
API price (in/out)
No hosted API
Modalities
text
Benchmark results
Bar shows position within the tracked field; marker = field best
best: Doubao-Seed-1.6 · 96.0%
best: Llama 3.1 405B · 96.8%
best: GPT-5 · 99.4%
best: Falcon-H1 34B · 83.6%
best: Doubao-1.5-Pro · 59.8%
Run it locally
VRAM @ Q4
4.5 GB
VRAM @ FP16
14 GB
Fits on (Q4)
RTX 3060 12GBRTX 4070 Ti 16GBRTX 3090 24GBRTX 4090 24GBRTX 5090 32GBM4 Pro 48GBM3 Max 128GBM3 Ultra 512GBA100 80GBH100 80GBH200 141GBB200 192GB
Throughput data unavailable.
Quantizations
GGUF
Fine-tune it
Conditional / customQLoRA6.4 GB1× RTX 3060 12GB
LoRA16.6 GB1× RTX 3090 24GB
Full fine-tune114.0 GB1× H200 141GB
QLoRA SFT on ~10k samples ≈ $4.10 (1× RTX 3060 12GB)
API price weights · each benchmark row carries its own source badge (see methodology)