Model

GLM-5.3 Flash Abliterated

Agentic all-rounder with strong terminal and tool-use scores.

Specs

Model id: glm-5.3-flash-abliterated
Context length: 262,144 tokens
Max output: 65,536 tokens (4,096 on trial accounts)
Input: Text only. Image, audio, video and file inputs are rejected.
Runs on: Dedicated GPUs operated by our infrastructure provider under a zero-data-retention configuration.
Precision: FP8 · Zero additional quantizationNative FP8 (block-FP8, e4m3): the precision Z.ai releases GLM-5.3 Flash in. Zero additional quantization: never compressed to INT8, INT4 or FP4.
Reasoning: Per request: reasoning_effort on Chat Completions, thinking on Messages, reasoning.effort on Responses.
Endpoints: /v1/chat/completions, /v1/messages, /v1/responses (stateless)
Base model: zai-org/GLM-5.3-Flash (MIT license)

Price

Compare providers

Input: $0.0977
Cached input: $0.0098
Output: $0.3421

USD per 1M tokens, billed from prepaid credit.

Uptime, last 30 days

Status page

100%

15 of 15 canary checks passed (one every 5 minutes).

Last canary: passed, time to first token 495 ms, total 526 ms.

Speed, last 24 h

—

Speed is measured, not guaranteed: the median output tokens per second after the first token, and the median time to first token (TTFT), over streamed requests in the last 24 h. "—" until a model has 20 of them.

Benchmarks

Benchmarks: base model, vendor-reported unless noted (zai-org/GLM-5.3-Flash)

Terminal-Bench 2.1: 84.3source
DeepSWE v1.1: 63.4source
HLE (with tools): 55.3source
GPQA-Diamond: 92.17NVIDIA BF16 run

Abliteration effect: MMLU (n=300) 83.3 → 82.7 (orcarouter build, reported by its publisher, source).

Quickstart

Create a key in the dashboard, then call glm-5.3-flash-abliterated from any OpenAI or Anthropic client.

curl https://api.abliteratedllm.dev/v1/chat/completions \
  -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "glm-5.3-flash-abliterated",
    "stream": true,
    "messages": [{"role": "user", "content": "Write the opening line of a noir detective story."}]
  }'