GLM-5.3 FlashX API

GLM-5.3 FlashX is Z.ai's speed-focused multimodal model for interactive coding and agent workflows. Compare token costs with Flash and integrate through EvoLink's unified API.

Z.aiText GenerationAvailable

from $0.371 / 1M input tokens

API docs

Multimodal inputImage · video · file inputReasoning always onChat + Messages

Production routeLive

Live

Context

1M context · 131K max output

Best For

Interactive coding agents, screenshot analysis, document workflows

Input

Text + images + video + files

Output

Text · JSON (structured output) · tool calls

Models Pricing Readme API

Choose GLM-5.3 FlashX

GLM-5.3 FlashX is Z.ai's speed-focused multimodal model for interactive coding and agent workflows. Compare token costs with Flash and integrate through EvoLink's unified API.

GLM-5.3 FlashX

Z.ai's speed-focused multimodal model

Selected

Model ID

glm-5.3-flashx

Best for

Interactive coding agents, screenshot analysis, document workflows

Input

$0.371 / 1M25.2 cr / 1M

Cache read

$0.075 / 1M5.1 cr / 1M

Output

$1.250 / 1M85 cr / 1M

All rates are per 1M tokens, shown in USD and credits, and reflect your account's current pricing.

GLM-5.3 FlashX pricing

Estimate what one GLM-5.3 FlashX request costs before you integrate. The calculator uses your account's current rates, with official pricing as a reference.

Request calculator

Enter the token mix for one request and the number of successful tool calls.

Input tokensCache read tokensOutput tokensWeb search

Estimated request cost

GLM-5.3 FlashX

USD$0.0008

Credits0.0518

Input tokens0.0252 cr

Cache read tokens0.0011 cr

Output tokens0.0255 cr

Minimum charge: 0.01 credits per request.

Budget guide

Approximate requests using the current token mix.

Add credits

$10

About 13127 requests

For quick testing

$50

About 65637 requests

For regular development

$100

About 131274 requests

For production evaluation

Server-side tool rates

Only successful server-side calls are billed per call; failed attempts have no tool fee, but tokens still apply.

Web search$0.010/ call0.68 cr / call

GLM-5.3 FlashX for interactive multimodal workflows

Choose FlashX when waiting for model output slows down a coding or agent workflow. Flash offers lower token prices; evaluate whether FlashX saves enough time on your own tasks to justify the difference.

GLM-5.3 FlashX is served on EvoLink under the model ID glm-5.3-flashx through Chat Completions · Responses · Anthropic Messages, with the same API key and balance you use for every other model. It offers a 1M context window and up to 131K output tokens, plus image input, long-document analysis, multi-step agents.

GLM-5.3 FlashX specs and capabilities

Numbers come from the EvoLink route configuration; capabilities are what the API exposes today.

Context window: 1M tokens
Max output: 131K tokens
Input: Text + images + video + files
Output: Text · JSON (structured output) · tool calls
Reasoning: Always-on reasoning
Tool use: Function calling with multi-step tool sequences
Prompt caching: Automatic cache reads at a lower rate
Server-side tools: Web search, billed per successful call
Protocols: Chat Completions · Responses · Anthropic Messages
Model ID: glm-5.3-flashx

What changes the FlashX selection decision

Balance the model's speed focus with input requirements, task quality, and the cost of a completed result.

A speed-focused option alongside Flash

Z.ai positions FlashX for faster inference. This does not establish a quality advantage over Flash or guarantee EvoLink latency. Compare both with the same prompts, reasoning settings, and output limits.

Multimodal evidence in a long context

Z.ai documents text, image, video, and file input with a 1M-token context window. Use the API documentation for the accepted request formats; the model returns text, while an agent's tools create files or perform actions.

Model choice through one gateway

Use EvoLink to compare model costs and keep integration in one API gateway. Keep Flash for work that meets your quality target at lower cost; evaluate FlashX where response time matters.

Where to evaluate GLM-5.3 FlashX

Start with tasks where a person or another tool is waiting for the next model response.

Interactive coding agents

Evaluate implementation, debugging, and tool-result interpretation in the same agent loop. Measure time to a working result, including tool execution and retries, rather than token speed alone.

Screenshot-driven development

Use interface screenshots to ground a code change or explain a visual defect. Check that the result matches the reference and that your chosen API route accepts the image format.

Document workflows with immediate feedback

Evaluate extraction and follow-up questions over documents with text and visual evidence. For offline classification or bulk processing where waiting is acceptable, compare the lower-priced Flash first.

Choose between FlashX, Flash, and GLM-5.3

EvoLink

Compare the current rates below. FlashX prioritizes speed; Flash has a lower token price. Measure task success and end-to-end time before changing models.

Model	GLM-5.3 FlashX	GLM-5.3	GLM-5.3 Flash
Input / output	$0.371 / $1.25	$1.4 / $4.4	$0.15 / $0.5
Context	1M	1M	1M
Caching	Cache reads	Cache reads	Cache reads
Best for	Interactive coding agents, screenshot analysis, document workflows	Evaluate the text-only flagship for demanding reasoning tasks. A multimodal request cannot be forwarded unchanged to a text-only model.	Lower-priced multimodal option for offline processing and cost-sensitive tasks. Compare FlashX when waiting for responses becomes a bottleneck; do not assume an accuracy gain.

GLM-5.3 FlashX

Input / output$0.371 / $1.25

Context1M

CachingCache reads

Best forInteractive coding agents, screenshot analysis, document workflows

GLM-5.3

Input / output$1.4 / $4.4

Context1M

CachingCache reads

Best forEvaluate the text-only flagship for demanding reasoning tasks. A multimodal request cannot be forwarded unchanged to a text-only model.

GLM-5.3 Flash

Input / output$0.15 / $0.5

Context1M

CachingCache reads

Best forLower-priced multimodal option for offline processing and cost-sensitive tasks. Compare FlashX when waiting for responses becomes a bottleneck; do not assume an accuracy gain.

Two ways to use GLM-5.3 FlashX: EvoLink API or Agent

Use the EvoLink API for product backends and batch jobs, or call GLM-5.3 FlashX from Codex, Claude, or Gemini for coding and analysis workflows. Both paths share the same EvoLink API key, balance, model ID, and request history.

Option 1

Integrate with the EvoLink API

Best for: product backends, batch jobs, automated pipelines

Send OpenAI-compatible Chat Completions (or Anthropic Messages) requests to EvoLink and control the model ID, system prompt, output budget, tools, and structured output.

1Create an EvoLink API key in the console
2Point your OpenAI or Anthropic SDK at the EvoLink base URL and select the model ID shown above
3Send one representative request and read the usage field for input, cached, and output tokens
4Set max_tokens and retries per task; keep tool-call IDs and results across turns

Create EvoLink API Key

Option 2

Call it with an Agent

Best for: coding, review, and analysis tasks in Codex, Claude, and Gemini

Give the Agent the task, the inputs to include, and the acceptance criteria. It assembles the request, calls GLM-5.3 FlashX through EvoLink, and returns the answer with token usage.

1Set EVOLINK_API_KEY in your local environment; never put it in code or a prompt
2Describe the task, the inputs to include, and the expected output format
3Ask the Agent to call GLM-5.3 FlashX through EvoLink and show the request before sending
4Let the Agent report the answer, token usage, and any error body

GLM-5.3 FlashX API code example and error handling

This example shows the shortest runnable request: an OpenAI-compatible Chat Completions call with a system prompt, a user message, and an output budget. Open the API tab for the complete parameter and response reference.

View complete API docs

cURL

curl -X POST https://api.evolink.ai/v1/chat/completions \
  -H "Authorization: Bearer $EVOLINK_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "glm-5.3-flashx",
    "messages": [
      { "role": "system", "content": "You extract structured data. Answer in JSON only." },
      { "role": "user", "content": "Extract the product, quantity, and unit price: 2 notebooks at $4.50 each." }
    ],
    "thinking": { "type": "enabled", "clear_thinking": false },
    "reasoning_effort": "low",
    "max_tokens": 4096,
    "temperature": 0.1
  }'

# Reasoning is always on for the 5.3 generation: do not send
# thinking.type "disabled". The same model ID works on /v1/messages
# (Anthropic Messages). The response includes choices[0].message and
# a usage object; image and video input is counted in prompt_tokens.

Invalid request or unsupported parameter

Check the model ID, the messages array, and parameter ranges against the API reference; remove fields this route does not support.

Authentication or balance issue

Check the Authorization bearer token and confirm the available balance in the console.

Context length exceeded

Prompt tokens exceed the GLM-5.3 FlashX context window. Trim or retrieve only the relevant evidence and reuse cached prefixes.

Rate limited (429)

Back off and retry with jitter; batch or queue requests instead of sending parallel bursts.

Content or tool call rejected

Review sensitive content, malformed tool-call arguments, and JSON schema mismatches before retrying.

Estimate input, output, and cache costs separately

Use this page's current rates and calculator for the same model ID. Check returned usage and billing on representative requests; a repeated prompt does not by itself prove a cache hit.

Measure the complete request

Record time to first token, output speed, total task time, retries, and cost per accepted result separately. A supplier's tokens-per-second figure does not include every step of your application.

GLM Model Family

Same API key and balance — switch tiers without changing your integration.

GLM-5.3

Z.ai flagship reasoning model

View model

GLM-5.2

The previous GLM flagship. Still relevant for clients that depend on disabling reasoning, which the 5.3 generation no longer allows.

View model

GLM-5.3 Flash

Z.ai multimodal high-volume model

View model

Other text models on EvoLink besides GLM-5.3 FlashX

DeepSeek V4 Flash

DeepSeek's high-volume route with a 1M context and very low cache-read pricing. The closest cost comparison for bulk text work.

View model

Gemini 3.7 Flash

Google's low-cost multimodal route — a useful cross-vendor baseline when image and document understanding drive the choice.

View model

Kimi K3

Moonshot’s long-context reasoning route for repository-scale coding and multi-document work.

View model

GPT-5.6

OpenAI’s tiered frontier family (Sol/Terra/Luna) for capability, latency, and cost-routing flexibility.

View model

GLM-5.3 FlashX guides and related reading

GLM-5.3 Flash vs GLM-5.3

Choose the family route by modality, task difficulty, and cost per accepted result — with a Flash-first escalation policy.

Read guide

One Gateway for 3 Coding CLIs

Config paths, environment variables, and a troubleshooting checklist for running coding CLIs through a single endpoint.

Read guide

GLM-5.3 FlashX API FAQ

What is GLM-5.3 FlashX?

FlashX is Z.ai's speed-focused option in the GLM-5.3 Flash family. It targets interactive coding and multimodal agent workflows; a higher token price does not establish higher answer quality.

What model ID does this EvoLink page use?

The model ID is glm-5.3-flashx. Use the API documentation linked on this page for endpoint-specific request formats and supported parameters.

How much does the GLM-5.3 FlashX API cost?

The pricing section shows input, output, and cache-read rates for this model. Use its calculator for your token mix, and verify actual usage and charges rather than applying another provider's price or promotion.

When should I choose FlashX instead of Flash?

Evaluate FlashX when response time limits an interactive workflow. Flash has a lower token price and may be preferable for offline batches. Compare task success, total time, retries, and cost with the same settings.

Does EvoLink guarantee 200 tokens/s?

No speed guarantee is made here. Z.ai reports 200 tokens/s for FlashX; that is a supplier claim, not an EvoLink benchmark or SLA. Measure first-token latency and total request time separately on your workload.

Which input and output types does FlashX support?

Z.ai lists text, images, video, and files as input, with text output. Check the linked API documentation for the formats supported by your route. Video or file input does not mean the API directly generates videos or Office files.

Can I disable reasoning?

Z.ai documents thinking.type as enabled only for FlashX. A lower reasoning_effort does not turn reasoning off. Z.ai recommends thinking.clear_thinking: false, nested inside the thinking object; check route support in the API documentation.

How should I budget for media and cached input?

Check usage and billing from representative requests, and apply the matching rates in the pricing section. Do not assume a fixed token count per image or a cache hit for every repeated prompt.

What context window does FlashX offer?

Z.ai documents a 1M-token context window and a maximum output of 128K tokens. Check the specifications and API documentation before choosing an output limit; input context and output allowance are different limits.

Can I use GLM-5.3 as a fallback?

Evaluate it for compatible text-only tasks that fail your acceptance checks. GLM-5.3 does not accept the same multimodal inputs, so do not forward image, video, or file requests unchanged. Define how inputs, cost, and retries will be handled first.