GLM 5.3
Z.ai released 2026-08-18 proprietary
Nothing is both better and cheaper.
Under a balanced workload, no other model in the catalogue scores higher and costs less. This model is on the value frontier.
Our take
editorial — not a measurementAA intelligence 60 and LMArena 1487 at max effort for $1.40/$4.40, with the lowest blended cost of any model in the Artificial Analysis top ten at $0.68 per million. On the intelligence-per-dollar axis this is arguably the best model in the dataset. Two caveats keep it out of a default recommendation: OpenRouter lists no Hugging Face repo for 5.3 (unlike GLM 5.2), so despite the family's open-weight history the weights status here is unconfirmed, and it is text-only with a 1M window. Benchmark it before assuming the score transfers to your task.
Strengths
- AA intelligence 60 at $1.40/$4.40 — best intelligence-per-dollar here
- Lowest blended cost ($0.68/M) in the AA top ten
- LMArena 1487 at max effort
- 1M context window
Weaknesses
- No Hugging Face repo listed — open-weight status unconfirmed for 5.3
- Text-only, no vision
- Licence unknown
- No output-speed or TTFT measurement available
Reach for it when
- Cost-constrained frontier-adjacent reasoning
- Text-only high-volume work
- Teams willing to benchmark before committing
Avoid it if
- You need multimodal input
- You need a confirmed open-weight licence
Every price dimension
| Input | $1.40/M |
|---|---|
| Output | $4.40/M |
| Cached input | $0.260/M |
Independent scores
| Intelligence | 59.5 |
|---|---|
| Coding | 74.8 |
| Agentic | 59.1 |
| LMArena Elo | 1476.9max |
Sources are listed separately rather than averaged. Across the 62 models both have scored they correlate at r = 0.806 — close agreement overall, but four models rank very differently between them, and a blended score would hide exactly those.
Capability
| Context window | 1.0M |
|---|---|
| Max output | 131K |
| Input modes | text |
| Tool use | yes |
| Reasoning | always on |
| Open weights | no |
Provenance
| Price source | openrouter.ai |
|---|---|
| Fetched | 2026-08-24 |
| Quality data | verified |
| Cross-checked | vendor page |
GLM-5.3 is a large-scale reasoning model from Z.ai, built for complex software engineering and long-horizon agent tasks. It supports text input and output with a 1M-token context window, and improves...