Autor: Jakub Rusinowski · Ostatnia aktualizacja: 11 września 2026
Model library → Devstral → Devstral-2 22B (Unverified Listing)
UNVERIFIED — retained only because this URL has been live and holds inbound links. Mistral shipped Devstral 2 in two sizes, Devstral 2 123B and Devstral Small 2 24B; no 22B Devstral 2 checkpoint appears under mistralai on Hugging Face or in the Devstral 2 announcement. The figures below were published here without sources. If you came looking for the consumer-GPU Devstral, you want Devstral Small 2 24B below — 68.0% on SWE-bench Verified at about 15 GB.
Devstral-2 22B (Unverified Listing) needs about 14 GB of VRAM at Q4_K_M — quantized weights plus framework overhead, before any KV cache. On Apple Silicon that figure comes out of unified memory.
| Parameters | 22 Billion (unverified) |
| Context window | 128,000 |
| Architecture | Dense Transformer |
| Provider | Mistral AI |
| Licence | Apache 2.0 |
| Specified at | Q4_K_M |
| System RAM | 32 GB |
| Record updated | 2026-09-11 |
Apache-2.0 — commercial use permitted. Commercial use permitted. No usage restrictions beyond attribution.
Modelled on a reference NVIDIA RTX 4090 (24 GB). Weights plus framework overhead only — this model publishes no architecture we can read, so no KV cache is included. A real session needs more; the figure is a floor, not a target. Speed figures are ESTIMATES from the memory-bandwidth roofline described on the methodology page, not benchmarks we ran — rows marked measured come from published or reader-submitted runs. VRAM here includes the KV cache, so it reads higher than the headline figure above, which does not.
| Quant | Bits/weight | Weights | VRAM needed | Est. speed | Fit on 24 GB |
|---|---|---|---|---|---|
| Q2_K | 2.63 | 7.2 GB | 8 GB | ~78 tok/s (est.) | Fits comfortably |
| Q3_K_M | 3.41 | 9.4 GB | 10.2 GB | ~64 tok/s (est.) | Fits comfortably |
| Q4_K_M | 4.83 | 13.3 GB | 14.1 GB | ~48 tok/s (est.) | Fits comfortably |
| Q5_K_M | 5.67 | 15.6 GB | 16.4 GB | ~42 tok/s (est.) | Fits comfortably |
| Q6_K | 6.56 | 18 GB | 18.8 GB | ~37 tok/s (est.) | Fits comfortably |
| Q8_0 | 8.50 | 23.4 GB | 24.2 GB | ~4 tok/s (est.) | Offloads to system RAM (slow) |
| F16 | 16.00 | 44 GB | 44.8 GB | ~2 tok/s (est.) | Offloads to system RAM (slow) |
Want to set your own context length and KV-cache quantization? Use the interactive VRAM calculator.
or compare on Vast.ai from $0.35/hr (typical low · varies)
As an Amazon Associate we earn from qualifying purchases. Cloud GPU links are referral links — we may earn a commission at no extra cost to you.
The cheapest catalogued GPU that runs Devstral-2 22B (Unverified Listing) is the AMD Radeon RX 9060 XT 16GB (16 GB).
Install Ollama, then run:
ollama run devstral-2Weights on Hugging Face: mistralai/Devstral-2-22B.
Best for: software engineering, coding, consumer gpu, local inference.
← All Devstral models | VRAM calculator | Check your own hardware