Models · Compare
GLM-4.6 vs Claude Sonnet 4.5
Rows highlighted in warm gray are where the models differ. Numbers carry their as-of date and primary source.
Specs
| Field | A: GLM-4.6 | B: Claude Sonnet 4.5 |
|---|---|---|
| Released | 2025-09-30 | 2025-09-29 |
| Developer | Zhipu AI | Anthropic |
| Openness | Open | Proprietary |
| License | MIT | Proprietary |
| OSI-approved | yes | no |
| Data released | no | no |
| Training code | no | no |
| Architecture | moe | unknown |
| Total params | 357B | — |
| Active params | — | — |
| Experts | — | — |
| Context window | 128K | — |
| Attention | unknown | unknown |
| Position enc. | unknown | unknown |
| Pretraining tokens | — | — |
| Post-training | sft, rlhf | rlhf, constitutional |
| Training hardware | — | — |
| $/M input | $0.60 | $3.00 |
| $/M output | $2.20 | $15.00 |
| Output tok/sec | 30.7 | 48.8 |
Benchmarks
Missing scores render as not reported; never inferred. Bold highlights the leader per benchmark.
General reasoning
| MMLU-Pro | 78.4 2026-05-21 | 86.0 2026-05-21 |
| GPQA-Diamond | 63.2 2026-05-21 | 72.7 2026-05-21 |
Code
| SWE-Bench Verified | — | 77.2 2025-09-29 |
| LiveCodeBench | 56.1 2026-05-21 | 59.0 2026-05-21 |
Math
| AIME 2025 | 44.3 2026-05-21 | 37.0 2026-05-21 |
Context · A
September 30 2025 refresh, 357B MoE with 200K context (up from 128K in 4.5) and 128K maximum output. Zhipu reports a 27 percent coding improvement over 4.5 and parity with Claude Sonnet 4 on 8 public benchmarks. MIT licensed.
Context · B
September 29 2025 incremental update of the Sonnet line at the same $3 / $15 price point as Sonnet 4. Shipped alongside Claude Code checkpoints, a native VS Code extension, context editing and memory tools for long-running agents, and the Claude Agent SDK. OSWorld score of 61.4% positioned it as the strongest computer-use model from Anthropic at launch.