Z.ai
GLM 5.3 Flash
Z.ai's 320B multimodal coding model pairs 18B active parameters with a one-million-token context window for code, tool use, and visual workflows.
Replaces GLM 4.7 Flash: input $0.0735 → $0.1575 per 1M (+114%), output $0.42 → $0.525 per 1M (+25%), context 200K → 1M.
Run this model
zai/glm-5.3-flash
const response = await fetch('https://api.minirouter.sh/v1/chat/completions', { method: 'POST', headers: { Authorization: `Bearer ${process.env.MINIROUTER_KEY}`, 'Content-Type': 'application/json', }, body: JSON.stringify({ model: 'zai/glm-5.3-flash', messages: [{ role: 'user', content: 'Why is the sky blue?' }], }),}) const data = await response.json()console.log(data.choices[0].message.content)Model overview
- Input / 1M
- $0.1575
- All-in MiniRouter rate
- Output / 1M
- $0.525
- All-in MiniRouter rate
- Context
- 1M
- tokens
- Max output
- 131K
- tokens
- Architecture
- 320B / 18B
- total / active
- Tested as
- ox-alpha
- OpenCode and OpenRouter
- Released
- 26 Aug 2026
- publisher release
The model behind ox-alpha
Fast enough to be the default.
Z.ai says it tested GLM-5.3-Flash anonymously as ox-alpha on OpenCode and OpenRouter, where it became the most popular model of the week.
The charts below are publisher-reported evidence, not MiniRouter-run benchmarks. The price shown by MiniRouter is the live all-in route price.
Cost / intelligence
On the reported Pareto frontier.
The launch chart places the discounted run at an Artificial Analysis Intelligence Index score of 57 and $0.045 per task.
- Index score
- 57
- Discounted cost / task
- $0.045
View chart data
| Field | Reported value | Context |
|---|---|---|
| Model | GLM-5.3-Flash | Anonymous test name: ox-alpha |
| Intelligence Index | 57 | Artificial Analysis v4.1.1 |
| Cost per task | $0.045 | Discounted run shown in chart |
| Comparison field | 59 models | Logarithmic cost axis |
Six evaluations
Coding and agent benchmarks
Z.ai reports gains over GLM-5.2 on every shown evaluation, including 63.4 vs 46.2 on DeepSWE and 48.8 vs 26.2 on AutomationBench.
View chart data
| Benchmark | GLM-5.3-Flash | GLM-5.2 | DeepSeek-V4-Vision-Exp | Claude Opus 4.8 | GPT-5.6 Terra | Gemini 3.7 Flash |
|---|---|---|---|---|---|---|
| Terminal Bench 2.1 | 84.3 | 81.0 | 83.9 | 85.0 | 87.4 | 85.8 |
| DeepSWE v1.1 | 63.4 | 46.2 | 59.3 | 58.0 | 69.6 | 65.3 |
| Agents' Last Exam | 26.3 | 20.4 | 27.3 | 27.0 | 28.0 | Not reported |
| AutomationBench v1.0.6 | 48.8 | 26.2 | 38.8 | 41.0 | 37.2 | 52.3 |
| HLE with tools | 55.3 | 54.7 | 55.1 | 57.9 | Not reported | Not reported |
| GDPval-AA v2 | 1773 | 1504 | 1675 | 1582 | 1571 | 1527 |
Z.ai Code Bench v1.0
Performance by reasoning effort
GLM-5.3-Flash leads GLM-5.2 at every reported effort level. At max effort it reaches 29.0% accuracy, near Claude Opus 4.8 at 29.5%.
View chart data
| Model and effort | Accuracy | Average output tokens per task |
|---|---|---|
| GLM-5.3 · low | 24.6% | about 49K |
| GLM-5.3 · high | 31.4% | about 51K |
| GLM-5.3 · max | 34.5% | about 75K |
| GLM-5.2 · non-thinking | 19.4% | about 44K |
| GLM-5.2 · high | 20.8% | about 49K |
| GLM-5.2 · max | 23.4% | about 95K |
| GLM-5.3-Flash · low | 21.4% | about 42K |
| GLM-5.3-Flash · high | 28.0% | about 71K |
| GLM-5.3-Flash · max | 29.0% | about 139K |
| Claude Fable 5 · low | 28.8% | about 32K |
| Claude Fable 5 · high | 35.6% | about 58K |
| Claude Fable 5 · max | 39.5% | about 115K |
| Claude Opus 4.8 · low | 21.1% | about 31K |
| Claude Opus 4.8 · high | 23.4% | about 51K |
| Claude Opus 4.8 · max | 29.5% | about 120K |
One route, standard clients.
Authenticate with a MiniRouter bearer key. Caller-provided upstream credentials are not used on this route.
Read model setup docs →https://api.minirouter.sh/v1zai/glm-5.3-flash/v1/chat/completions/v1/messages/v1/responsesmax_tokens, temperature, stop, tools, tool_choice, reasoning, include_reasoning
Configured provider
Priority follows the live catalog order.
Input: text, image · Output: text
| Provider | Priority | Input / 1M | Output / 1M | Parameters | Access |
|---|---|---|---|---|---|
vercel | 1 | $0.1575 | $0.525 | max_tokens, temperature, stop, tools, tool_choice, reasoning, include_reasoning | MiniRouter balance |