GLM 5.3 Flash

New
Beta
Turbo

Multimodal coding agents with 1M context

Chat
Multi-Modal
Thinking
JSON Mode
Tool Calling

Pricing

Run instantly. Pay only for what you use.

Input
$0.10/ M
Output
$0.35/ M
Cache read
$0.0175/ M

About GLM 5.3 Flash

GLM 5.3 Flash is Z.ai's first natively multimodal GLM 5 model, released under MIT: a 320B-parameter MoE that activates just 18B per token, with a 1M-token context window, up to 128K output tokens, and image and video input alongside text.

Its hybrid attention — linear KDA layers plus NoPE sparse MLA — runs on roughly 3× less attention compute with a 4.4× smaller KV cache than the flagship GLM 5.3, so whole-repo prompts, long video transcripts, and heavily cached agent context stay fast and cheap at full 1M context.

The result is GLM 5.2-class agentic work at a fraction of the cost: 84.3 on Terminal-Bench 2.1 and 63.4 on DeepSWE 1.1 at maximum reasoning effort — up from 46.2 for GLM 5.2 — with strict function calling and JSON output in every mode.

18Bactive of 320B parameters

Sparse routing keeps per-token compute near a small model while the full expert pool stays available.

84.3on Terminal-Bench 2.1

Within 0.7 points of Claude Opus 4.8 on end-to-end terminal tasks — from MIT-licensed open weights.

97%avg. cache hit rate on Entrim

Measured on coding-agent workloads — repeated context bills at the cache-read rate.

Key capabilities

GLM 5.3 Flash is strongest where agents run long, look at screens, and have to stay cheap — coding and automation loops that mix code, screenshots, and recordings in one context.

Native image and video input

Trained on a 30T-token multimodal corpus, it reads screenshots, diagrams, and screen recordings directly — no separate vision model in the loop.

1M tokens without the memory bill

Linear KDA layers plus NoPE sparse MLA cut attention compute by about 3× and the KV cache by 4.4× versus the flagship GLM 5.3 — whole-repo prompts run at full context.

Selectable reasoning effort

Low, high, and max efforts set thinking depth per request — deep reasoning for the hard steps, shorter traces where a request does not need them.

Strict tool calling and JSON output

Function calling and structured JSON output in every mode, with 48.8 on AutomationBench — nearly double GLM 5.2 — for multi-step tool use.

Where this model fits

Best for
  • Long-running coding agents

    Plans, edits, and verifies across long tool-call chains — 84.3 on Terminal-Bench 2.1 at maximum reasoning effort.

  • Agents grounded in screenshots and video

    Native image and video input means UI screenshots, diagrams, and screen recordings go straight into the prompt without a separate vision model.

  • Whole-repo and long-document analysis

    The 1M-token window holds a full repository or document set in one prompt, and the 4.4× smaller KV cache keeps it affordable.

  • Browser and workflow automation

    48.8 on AutomationBench — nearly double GLM 5.2 — for multi-step tool use across real applications.

Avoid for
  • Latency-critical short replies

    Thinking cannot be switched off — even low reasoning effort spends tokens before the reply. Use DeepSeek V4 Flash in non-think mode for fast conversational turns.

  • Production traffic during the beta

    GLM 5.3 Flash is served in NVFP4 as a beta on Entrim. Validate outputs against your own evals before moving production traffic, and keep DeepSeek V4 Flash as the fallback route.

  • Audio input or speech understanding

    Inputs are text, image, and video only. Transcribe audio upstream before sending it to the model.

What Turbo means

Turbo is Entrim's NVFP4 serving mode for GLM 5.3 Flash — the same open weights, quantized only where it is safe, for higher throughput at a lower price.

Only the routed-expert weights — about 97% of the model’s parameters — are stored in NVFP4. Attention layers, routers, embeddings, the vision tower, and all activations stay in higher precision, so 4-bit noise never touches the layers most sensitive to it.

That cuts the weight footprint by about 70%, which is what makes the 1M-token context practical to serve at Flash pricing.

FP8 vs NVFP4, measured by NVIDIA on GLM 5.2

BenchmarkFP8NVFP4Δ
GPQA Diamond89.5289.39−0.13
SciCode49.8549.04−0.81
IFBench74.9575.81+0.86
AA-LCR (long context)69.3870.13+0.75
τ²-Bench Telecom97.9098.25+0.35

The closest published FP8 vs NVFP4 comparison in the GLM family — measured by NVIDIA on GLM 5.2, not on GLM 5.3 Flash. Entrim’s own GLM 5.3 Flash evals will replace this table.

GLM 5.3 Flash Turbo is in beta. Validate outputs against your own evals before moving production traffic, and keep DeepSeek V4 Flash as the fallback route.

Performance and
benchmarks

Entrim performance for GLM 5.3 Flash under production load in the past few days.

Throughput (tokens / sec)

avg. 110tokens/s

Time to first token (TTFT)

avg. 1520ms

Public model benchmarks

84.3Terminal-Bench 2.1

Agentic coding and shell tasks run end-to-end in a real terminal.

63.4DeepSWE 1.1

Multi-step software engineering fixes — up from 46.2 for GLM 5.2.

48.8AutomationBench

Tool-driven automation across real applications and workflows.

Run GLM 5.3 Flash in minutes

Test this model with $10 free credit. Use Entrim's OpenAI-compatible API to call GLM 5.3 Flash, compare output quality, latency, and request cost, then decide if it fits your workload.

Code examples

curl -X POST "https://api.entrim.ai/v1/chat/completions" \  -H "Content-Type: application/json" \  -H "Authorization: Bearer $ENTRIM_API_KEY" \  -d '{    "model": "zai-org/GLM-5.3-Flash",    "messages": [      {        "role": "user",        "content": "Why do they call it a building if it is already built?"      }    ]  }'

SDKs & docs

Use familiar SDK patterns with quickstart examples for common production setups.

OpenAI-compatible

Keep the request format your team already knows.

No cold starts

Production requests are served without model spin-up delays.

Not sure this is the right model?

Explore related models that may be a better fit for your workload.

High-throughput agentic work with 1M context

1M
FP8
Thinking
Tool Calling
Input / M
$0.09
Output / M
$0.17
Cached / M
$0.015

Frontier-parity coding agents with 1M context

1M
FP8
Thinking
Tool Calling
Input / M
$0.9
Output / M
$3.0
Cached / M
$0.15

Dense multimodal coding with thinking control

262K
FP8
Vision
Multimodal
Video
Input / M
$0.10
Output / M
$0.40
Cached / M
$0.04

Run GLM 5.3 Flash with the Most Cost-Effective LLM Inference

Use $10 free credits to test prompts, token sizes, latency, throughput, output quality, and request cost before moving traffic.
All services are online

© 2026. Entrim. All Rights Reserved.

Privacy policyTerms of service