GLM-5.3
The open model people actually ship coding agents on. Same 744B base as GLM-5.2 with heavy agentic post-training; drops into Claude Code harnesses and holds its own on SWE-Bench Pro V2 and MCP Atlas at a third of frontier prices.
Key facts
- Score
- 81 / 100, rank #5Evals84Price47People85Fit100
- Price
- $1.40 in · $4.40 out per 1M tokens (Z.ai Pricing)
- Blended price
- $2.15 per 1M tokens (3 input : 1 output)
- Context window
- 1M
- Open weights
- Yes
- Released
- Aug 14, 2026
- Best for
- Agentic coding in Claude Code / Pi-style harnesses, Security research and code auditing, Self-hosted frontier-ish reasoning
GLM-5.3 costs $1.40 per million input tokens and $4.40 per million output tokens on Z.ai's list price (Z.ai Pricing, read Oct 11, 2026).
Independent evals
| Benchmark | Result | Who ran it | Read |
|---|---|---|---|
| Artificial Analysis Intelligence Index | 45 | GLM-5.3 - Artificial Analysis | Oct 10, 2026 |
| Arena Text score | 1478 (rank 27) | Text Arena Leaderboard | Oct 10, 2026 |
| Scale SWE-Bench Pro V2 (mini-swe-agent) | 84.30% | SWE-Bench Pro V2 Leaderboard - Scale Labs | Oct 10, 2026 |
| Scale MCP Atlas | 84.20 | MCP Atlas Leaderboard - Scale Labs | Oct 10, 2026 |
| Arena WebDev score | 1622 (rank 22) | WebDev Arena Leaderboard | Oct 10, 2026 |
What practitioners say
One of the biggest HN launches of the year (1,171 points, 584 comments). Practitioners report it 'fits Claude Code as its own, zero issues', that '$5 in tokens' of Claude work costs '$0.50' on GLM via Pi, and that it 'routinely finds bugs missed by Fable and Sol' in audits. The cyber angle was the controversy: users bragged about adapting kernel exploits that 'Claude and Opus outright refused', and others flagged the risk of nerfed public weights.
Caveats
- License is muddled: AA lists a custom 'GLM-5.3 License' with commercial restrictions while Arena lists MIT. Read it before you self-host.
- Permissive on offensive-security tasks; that is a feature for red teams and a governance problem for everyone else.
- Open weights shipped ~2 weeks after the API; HN worried the public weights were safety-tuned differently.
Questions people ask
How much does GLM-5.3 cost?
GLM-5.3 costs $1.40 per million input tokens and $4.40 per million output tokens on Z.ai's list price (Z.ai Pricing, read Oct 11, 2026).
How good is GLM-5.3?
GLM-5.3 ranks #5 of 26 on the Models at Work leaderboard with a score of 81 out of 100 (October 2026 edition). The open model people actually ship coding agents on. Same 744B base as GLM-5.2 with heavy agentic post-training; drops into Claude Code harnesses and holds its own on SWE-Bench Pro V2 and MCP Atlas at a third of frontier prices.
What is GLM-5.3 best for?
GLM-5.3 is best for agentic coding in claude code / pi-style harnesses, security research and code auditing, self-hosted frontier-ish reasoning.
What are GLM-5.3's benchmark scores?
Artificial Analysis Intelligence Index: 45 (GLM-5.3 - Artificial Analysis); Arena Text score: 1478 (rank 27) (Text Arena Leaderboard); Scale SWE-Bench Pro V2 (mini-swe-agent): 84.30% (SWE-Bench Pro V2 Leaderboard - Scale Labs); Scale MCP Atlas: 84.20 (MCP Atlas Leaderboard - Scale Labs); Arena WebDev score: 1622 (rank 22) (WebDev Arena Leaderboard).
What is GLM-5.3's context window?
GLM-5.3 has a 1M token context window according to Z.ai.
Is GLM-5.3 open weights?
Yes. GLM-5.3 is released with open weights, so it can be self-hosted.
Compare GLM-5.3
Methodology: Score is 0 to 100 and computed, not typed: 55% independent evals (Artificial Analysis, Arena, Scale SEAL, Epoch, each scaled against its natural floor and the best score in this table, then averaged), 15% blended price on a fixed log scale ($0.05 per million is 100, $60 is 0), 15% practitioner sentiment from Signals and community threads, 15% operational fit (context window, open weights). A missing component drops out and the rest are reweighted. A model with no independent eval yet is provisional and ranks below every measured one. Within three points is a tie. Vendor-published figures are labelled and never counted.