All models › deepseek
Arena score1425#104 of 416 in Chat, paid models included
Size685.4 B51.9 B active per token (mixture of experts)
4-bit download415 GBweights only, Q4_K_M
Max context160Ktokens, from the model config
Where to use it for free
- The lab's own chat: chat.deepseek.com (free with daily limits)
- Download and run it at home: official weights · GGUF files for Ollama and LM Studio
Arena scores
| Arena | Score | Rank | Votes |
|---|---|---|---|
| Chat | 1425 | #104 of 416 | 9K |
Rank among all models, paid ones included. Source: Arena leaderboard dataset.
Speed on popular hardware
| Hardware | GPU memory | Fits? | Speed |
|---|---|---|---|
| RTX 3060 12 GB | 12 GB | Does not fit | — |
| RTX 4060 Ti 16 GB | 16 GB | Does not fit | — |
| RTX 3090 | 24 GB | Does not fit | — |
| RTX 4090 | 24 GB | Does not fit | — |
| RTX 5090 | 32 GB | Does not fit | — |
| RX 7900 XTX | 24 GB | Does not fit | — |
| 2 × RTX 3090 | 48 GB | Does not fit | — |
| 4 × RTX 3090 | 96 GB | Does not fit | — |
| Mac M4 Max 128 GB | 96 GB | Does not fit | — |
| Ryzen AI Max+ 395 128 GB | 96 GB | Does not fit | — |
| 8 × H100 SXM | 640 GB | Fits in GPU | 57 tok/s |
Estimate at 4-bit with 32K context, 64 GB of DDR5 for offload. Speed ≈ memory bandwidth × 70% ÷ data read per token; real numbers vary by about ±30%.
GPU memory needed
| Level | 8K context | 32K context | 128K context | Quality vs full |
|---|---|---|---|---|
| Q3_K_M | 344 GB | 346 GB | 353 GB | +4.7% noticeable loss |
| Q4_K_M | 417 GB | 418 GB | 425 GB | +1.2% very small loss |
| Q5_K_M | 488 GB | 490 GB | 497 GB | +0.4% near-lossless |
| Q6_K | 563 GB | 565 GB | 572 GB | +0.2% near-lossless |
| Q8_0 | 729 GB | 731 GB | 738 GB | ≈ 0% same as full |
| F16 | 1372 GB | 1374 GB | 1381 GB | ≈ 0% same as full |
Weights + context cache (from the model's architecture) + runtime. Quality: increase in perplexity against the full model, from llama.cpp tests, scaled by model size. Simulate your own setup.
Chat score over time
1400145015001550
MayJunJulAugSeptOct
deepseek-v3.2-exp-thinking 1425Best paid model
Arena score from 17 Apr 2026 to 15 Oct 2026.
Similar free models in Chat
- DeepSeek-V3.2 (Thinking) 1424 · 685.4 B
- deepseek-v4-flash-high-preview 1424
- longcat-flash-chat 1423 · 561.9 B
- Mistral Large 3 1428
- MiMo-V2.5 1428 · 310.8 B
- deepseek-r1-0528 1428 · 684.5 B