Side by side
See the difference.
Scores, capabilities and costs for the models you’re considering.
Scroll to compare models →
| 3 models | zai-orgGLM-5.3-FlashRemove × | stepfun-aiStep-5-PreviewRemove × | openaiOpenAI: GPT-6 LunaRemove × |
|---|---|---|---|
| What’s different | GLM-5.3-Flash is a native multimodal model from Z.ai. It is suited for efficient coding and long-horizon agent tasks. Its hybrid sparse and linear attention architecture maintains accurate long-context… | 600B total and 27B active parameters; accepts images and video, with FP8 and BF16 weights for deployment. | Can turn reasoning off for high-volume requests; Chat Completions tool calls require reasoning_effort=none. |
| Date | Aug 26, 2026Released | Sep 18, 2026Released | Sep 22, 2026Released |
API prices USD · provider configuration shown below | |||
| Listed rate | $0.06 in / $1.80 out/ 1M tokensReka · via OpenRouter · all routes ↗reka | $1.00 in / $2.70 out/ 1M tokensDeepInfra · all routes ↗ | $0.10 in / $0.50 out/ 1M tokensAzure · via OpenRouter · all routes ↗azure · prompt ≤ 272000 tokens |
| Provider endpoints | 46 endpoints → | 6 endpoints → | 9 endpoints → |
Benchmarks Each result retains its evaluation configuration | |||
| AA Intelligence ↗4.3.2 | 41.81 ↗Effort: maxIndependent result | 43.73 ↗Effort not specifiedIndependent result | 38.12 ↗Effort: max · MaxIndependent result |
| AA-Briefcase ↗1.1 | 1,452.63 ↗Effort: maxIndependent result | 1,424.32 ↗Effort not specifiedIndependent result | 1,334.83 ↗Effort: max · MaxIndependent result |
| GDPval-AA ↗2.1 | 1,642.18 ↗Effort: maxIndependent result | 1,586.63 ↗Effort not specifiedIndependent result | 1,431.86 ↗Effort: max · MaxIndependent result |
| AutomationBench-AA ↗AA Index 4.3.2 collection | 60.37% ↗Effort: maxIndependent result | 50.96% ↗Effort not specifiedIndependent result | 53.17% ↗Effort: max · MaxIndependent result |
| Terminal-Bench 4.0 · AA ↗ | 32.83% ↗Effort: maxHarness unreportedIndependent result | 33.33% ↗Effort not specifiedHarness unreportedIndependent result | 12.63% ↗Effort: max · MaxHarness unreportedIndependent result |
| SciCode ↗AA Index 4.3.2 collection | 51.62% ↗AA · Under reviewEffort: maxIndependent result | 58.91% ↗AA · Under reviewEffort not specifiedIndependent result | 54.63% ↗AA · Under reviewEffort: max · MaxIndependent result |
| Humanity’s Last Exam ↗AA Index 4.3.2 collection | 39.85% ↗Effort: maxIndependent result | 46.48% ↗Effort not specifiedIndependent result | 38.51% ↗Effort: max · MaxIndependent result |
| GDP.pdf ↗AA Index 4.3.2 collection | 15.4% ↗Effort: maxIndependent result | 14.8% ↗Effort not specifiedIndependent result | 22.8% ↗Effort: max · MaxIndependent result |
| CritPt ↗AA Index 4.3.2 collection | 15.43% ↗AA · Under reviewEffort: maxIndependent result | 20.86% ↗AA · Under reviewEffort not specifiedIndependent result | 19.43% ↗AA · Under reviewEffort: max · MaxIndependent result |
| AA-Omniscience ↗AA Index 4.3.2 collection | 7.47 ↗Effort: maxIndependent result | 16.38 ↗Effort not specifiedIndependent result | 0.65 ↗Effort: max · MaxIndependent result |
| AA-LCR ↗1.1 | 80% ↗Effort: maxIndependent result | 88.33% ↗Effort not specifiedIndependent result | 83.33% ↗Effort: max · MaxIndependent result |
| LiveBench ↗2.0 · 2026-10-14 (dataset 2026-09-02) | 70.61% ↗Effort: max · single run · equal mean of 7 suitesIndependent result | 62.01% ↗Effort: max · single run · equal mean of 7 suitesIndependent result | 69.4% ↗Effort: max · single run · equal mean of 7 suitesIndependent result |
| LiveBench Design ↗2.0 · 2026-10-14 (dataset 2026-09-02) | 77.2% ↗Effort: max · single runIndependent result | 76.9% ↗Effort: max · single runIndependent result | 82.5% ↗Effort: max · single runIndependent result |
| LiveBench Research ↗2.0 · 2026-10-14 (dataset 2026-09-02) | 70.1% ↗Effort: max · single runIndependent result | 72.8% ↗Effort: max · single runIndependent result | 43.5% ↗Effort: max · single runIndependent result |
| LiveBench Browser Use ↗2.0 · 2026-10-14 (dataset 2026-09-02) | 84% ↗Effort: max · single runIndependent result | 24% ↗Effort: max · single runIndependent result | 72% ↗Effort: max · single runIndependent result |
| LiveBench Coding ↗2.0 · 2026-10-14 (dataset 2026-09-02) | 80.6% ↗Effort: max · single runIndependent result | 85.6% ↗Effort: max · single runIndependent result | 84.5% ↗Effort: max · single runIndependent result |
| LiveBench Agentic ↗2.0 · 2026-10-14 (dataset 2026-09-02) | 28.4% ↗Effort: max · single runIndependent result | 27.5% ↗Effort: max · single runIndependent result | 48.1% ↗Effort: max · single runIndependent result |
| LiveBench Data Analysis ↗2.0 · 2026-10-14 (dataset 2026-09-02) | 76.4% ↗Effort: max · single runIndependent result | 63.5% ↗Effort: max · single runIndependent result | 73.4% ↗Effort: max · single runIndependent result |
| LiveBench Reasoning ↗2.0 · 2026-10-14 (dataset 2026-09-02) | 77.6% ↗Effort: max · single runIndependent result | 83.8% ↗Effort: max · single runIndependent result | 81.8% ↗Effort: max · single runIndependent result |
Specifications | |||
| Context window | 1M | 1M | 1.1M |
| Max output | 943.7K | 64K | 128K |
| Parameters | — | — | — |
| Tool calling | Supported | Supported | Supported |
| Structured output | Supported | Supported | Supported |
| Image input | Supported | Supported | Supported |
| Video input | Supported | Supported | Not reported |
| Reasoning | Supported | Supported | Supported |
| Prompt caching | Supported | Supported | Supported |
| Open weights | Supported | Supported | Not reported |