GLM-4.5-Air vs Devstral-2-123B-Instruct-2512
Specs, VRAM requirements and download trends — updated 24 October 2026.
Specification comparison
Differences are highlighted; identical values are muted.
| Specification | GLM-4.5-Air | Devstral-2-123B-Instruct-2512 |
|---|---|---|
| Deployment | Open weights · also via API | Open weights · also via API |
| Parameters | 110.5 B | 125.0 B |
| Architecture | Glm4MoeForCausalLM | Ministral3ForCausalLM |
| Context length | 131,072 | 262,144 |
| Licence | mit | other |
| Commercial use | Allowed | Unknown |
| Languages | 2 | — |
| Downloads 30d | 850,129 | 5,315 |
| Downloads all time | 5.5 M | 353 K |
| Quantizations on the Hub | SAFETENSORS | SAFETENSORS |
| Gated | No | No |
| First seen on the Hub | 2025-07-20 | 2025-11-28 |
| API input price / 1M tokens | $0.13 | $0.40 |
| API output price / 1M tokens | $0.85 | $2.00 |
| Est. cost per 1M input + 1M output tokens | $0.98 | $2.40 |
| Cached input / 1M tokens | $0.025 | $0.04 |
| AA Intelligence Index | 11.1 | 8.6 |
| AA Coding Index | — | 31.3 |
| AA Agentic Index | — | 2.3 |
| Max output tokens | 98,304 | 209,715 |
| API providers | 3 | 1 |
| Tool calling | Yes | Yes |
| Reasoning mode | Yes | No |
API cost
Estimated cost of 1M input + 1M output tokens via OpenRouter
GLM-4.5-Air
$0.98
Devstral-2-123B-Instruct-2512
$2.40
GLM-4.5-Air is 2.4× cheaper at list price.
Download trend
Daily snapshots, last 89 days (28 Jul – 24 Oct)
850 K5 K
GLM-4.5-Air
Devstral-2-123B-Instruct-2512
VRAM side by side
On RTX 4090 · 24 GB · 8K context unless noted
GLM-4.5-Air · fp16 @ 128K ctx
508.7 GB / 24 GB
Devstral-2-123B-Instruct-2512 · fp16 @ 256K ctx
882.8 GB / 24 GB