GLM 5.3 Flash
Zhipu AI
glm-5.3-flashThe first natively multimodal model in the GLM 5 series, with hybrid sparse and linear attention.
- Reasoning
- Vision
- Tool use
- JSON mode
- Fixed price
- Input
- $0.1000
- Cache hit $0.0200
- Output
- $0.3000
- Per 1M tokens
- Context
- 1M
- 128k max output
- First token
- 1.71 s
- Typical, last 7 days
- Throughput
- 159 tok/s
- Typical, last 7 days
Specifications
- Developer
- Zhipu AI
- Architecture
- Mixture of experts, 320B total and 18B active
- Weights
- FP8
- Context length
- 1,048,576 tokens
- Max output
- 128,000 tokens
- Modalities
- Text and images in, text out
- Reasoning effort
- low, high, max; default max
- Model id
- glm-5.3-flash
- Pricing
- Fixed price
- Weights published
- zai-org/GLM-5.3-Flash
max_tokens above the max output is clamped down to it.
Prices
USD per 1M tokens- Cache-hit input
- $0.0200
- Cache-miss input
- $0.1000
- Output
- $0.3000
A typical request
A 10,000-token prompt, half of it cached, with a 1,000-token answer: $0.0009
GET /v1/models carries the prices in force.
Performance
last 7 days · OpenFill trafficTime to the first token, generation speed and cache hits on this model's own requests.
- First token
- 1.71 s
- p95 4.23 s
- Throughput
- 159 tok/s
- Slowest 5% under 96.6 tok/s
- Cache hits
- 93%
- Of prompt tokens
- Requests
- 153
- 100% streamed
| Prompt | First token | Throughput | Cache hits |
|---|---|---|---|
| Under 1k | 713 ms | 147 tok/s | 0% |
| 1k to 4k | Pending | Pending | Pending |
| 4k to 16k | Pending | Pending | Pending |
| 16k to 64k | Pending | Pending | Pending |
| 64k to 256k | 2.26 s | 177 tok/s | 97% |
| 256k and over | Pending | Pending | Pending |
Typical requestFirst token at p95
Uptime, last 24 hours
Each check is a chat completion through the public API, run the way the models board describes.
Uptime
Pending
Checks
0
Failures
0
Latency
Pending
Queue wait
Pending
Checks begin once capacity is serving this model.
24h agoNow
All checks passedFailed, drawn in proportionAwaiting capacity
Model
glm-5.3-flash Base URL https://api.openfill.ai/v1Set OPENFILL_API_KEY to a key from API keys before running an example. A key starts with of_live_ and is shown once.
import osfrom openai import OpenAI client = OpenAI( base_url=( "https://api.openfill.ai/v1" ), api_key=os.environ[ "OPENFILL_API_KEY" ],) r = client.chat.completions.create( model="glm-5.3-flash", messages=[ {"role": "user", "content": "hi"} ],)print(r.choices[0].message.content)print(r.usage)import OpenAI from "openai"; const client = new OpenAI({ baseURL: "https://api.openfill.ai/v1", apiKey: process.env.OPENFILL_API_KEY,}); const r = await client.chat.completions.create({ model: "glm-5.3-flash", messages: [{ role: "user", content: "hi" }],});console.log(r.choices[0].message.content);console.log(r.usage);curl "https://api.openfill.ai/v1/chat/completions" \ -H "Authorization: Bearer $OPENFILL_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "model": "glm-5.3-flash", "messages": [ {"role": "user", "content": "hi"} ] }'