Z.ai (GLM)

GLM 5.3 Flash

z-ai/glm-5.3-flash

Model weightsReleased Aug 26, 2026MIT

ThinkingToolsVisionopen weights

GLM-5.3-Flash is a native multimodal model from Z.ai. It is suited for efficient coding and long-horizon agent tasks. Its hybrid sparse and linear attention architecture maintains accurate long-context behavior while...

Run in ChatOSS Cloud Plans

Cost / 1M tokens

Very low usage

$0.04

input

$0.03

cached input

$0.078

output

Context Window

Up to 131K max output

Architecture

18B active MoE · 1M context

text + image + video → text

Quickstart & API

z-ai/glm-5.3-flash

Open the model picker in ChatOSS Desktop (Code, Kanban, or Chat) and select or paste this model ID.

Best for & Strengths

Key Strengths

18B active parametersMultimodal vision + videoMIT licenseHigh efficiency ($0.15/M)

Recommended Workloads in ChatOSS

Fast terminal agents
Multimodal debugging
Everyday coding workflows

Available in ChatOSS Code, Kanban Orchestrate, and Chat modes via Plus, Max, or BYOK.

Verified Evaluations

Benchmarks & Frontier Comparison

How GLM 5.3 Flash compares against leading open-weight and proprietary frontier models across independent agentic coding and reasoning suites.

View full Top 40 Leaderboard →

GLM 5.3 FlashDeepSeek V4.1 FlashGLM 5.3Kimi K3.1 Pro

Benchmark (Metric)	GLM 5.3 Flash	DeepSeek V4.1 Flash	GLM 5.3	Kimi K3.1 Pro	Claude Opus 5.5
Agentic & Software Engineering
ChatOSS Agentic Coding Index (%)	36.9%	63.2%	58.3%	65%	74.1%
SWE-Bench Verified (%)	92%	91.5%	95.4%	94.8%	97%
DeepSWE v1.1 (Pass@1 %)	63.4%	77%	69%	73.5%	74.2%
Vibe Code Bench v1.1 (%)	30.76%	87.7%	78.1%	88.4%	90.3%
Terminal-Bench 4.0 (%)	32.8%	42.3%	42%	46.2%	65.2%
Code Migration (%)	20.52%	45.6%	44.2%	52%	66.7%
Intelligence & Token Economics
Artificial Analysis Intelligence Index (0–100)	27.8	39.5	44.8	55.1	57.6
Input Price (/ 1M tokens)	$0.04	$0.15	$1.40	$2.00	$4.00
Output Price (/ 1M tokens)	$0.078	$0.60	$4.40	$10.00	$20.00

ChatOSS Agentic Coding Index

ChatOSS Benchmark Audit · 4-benchmark composite mean

36.9%

SWE-Bench Verified

Vals AI · Claude Code

92%

DeepSWE v1.1

DeepSWE / Official Audit · Mini-SWE-Agent

63.4%

Vibe Code Bench v1.1

Vals.ai · OpenHands

30.76%

Terminal-Bench 4.0

Vals.ai / Artificial Analysis · Terminus 2

32.8%

Terminal-Bench 2.1

Vals.ai · Claude Code

84.3%

Code Migration

Vals.ai · Vals evaluation harness

20.52%

AutomationBench

Artificial Analysis v4.3

60.4%

Artificial Analysis Intelligence Index

Artificial Analysis

27.8

GPQA Diamond (Pass@1)

Benchmark Audit

86.4%

Humanity's Last Exam (Pass@1)

Artificial Analysis / Audit

39.9%

Design Arena: 3D

Design Arena · Rank #20 (55.6% win)

1300 Elo

Design Arena: ASCIIART

Design Arena · Rank #14 (53.7% win)

1266 Elo

Design Arena: Code Generation

Design Arena · Rank #26 (49% win)

1273 Elo

Design Arena: Data Visualization

Design Arena · Rank #32 (50.3% win)

1259 Elo

Active Provider Endpoints (42)

Upstream inference providers serving GLM 5.3 Flash, with per-provider pricing, context limits, quantization, and 24-hour uptime.

Provider	Input / 1M	Output / 1M	Cache Read / 1M	Context	Max Output	Quantization	Uptime (24h)
Relacecache relace/fp4	$0.04	$0.078	$0.03	1M	131K	fp4	100.0%
Wafercache wafer/us	$0.042	$0.184	$0.041	1M	944K	nvfp4	100.0%
PrimeIntellectcache primeintellect	$0.052	$0.25	$0.011	1M	131K	—	100.0%
OpenInferencecache open-inference/fp4	$0.06	$0.30	$0.03	1M	944K	fp4	99.2%
Darkbloomcache darkbloom/fp4	$0.065	$0.18	$0.012	262K	66K	fp4	97.3%
Wafercache wafer/nvfp4	$0.065	$0.24	$0.064	1M	944K	nvfp4	100.0%
Tenstorrentcache tenstorrent/fp8	$0.075	$0.25	$0.015	262K	131K	fp8	99.3%
GMICloudcache gmicloud/fp8	$0.075	$0.25	$0.015	1M	944K	fp8	100.0%
Near AIcache near-ai/fp8	$0.08	$0.50	$0.012	1M	944K	fp8	100.0%
Novitacache novita/fp8	$0.084	$0.28	$0.017	1M	131K	fp8	100.0%
StreamLakecache streamlake/fp8	$0.088	$0.295	$0.018	1M	128K	fp8	97.8%
Morphcache morph/fp8	$0.093	$0.122	$0.04	1M	944K	fp8	100.0%
DekaLLMcache dekallm	$0.10	$0.50	$0.03	1M	944K	—	97.0%
InferenceNetcache inference-net/us	$0.108	$0.35	$0.035	1M	262K	fp4	100.0%
InferenceNetcache inference-net/fp4	$0.108	$0.35	$0.035	1M	262K	fp4	100.0%
DigitalOceancache digitalocean	$0.113	$0.375	$0.03	1M	131K	—	99.8%
Rekacache reka	$0.12	$0.30	$0.03	1M	944K	—	99.9%
CoreWeavecache coreweave/nvfp4	$0.12	$0.38	$0.02	1M	944K	nvfp4	67.6%
NextBitcache nextbit/fp4	$0.12	$0.45	$0.03	1M	128K	fp4	100.0%
Particle AIcache particle-ai/nvfp4	$0.12	$0.50	$0.04	1M	131K	nvfp4	99.3%
Makoracache makora/fp4	$0.14	$0.24	$0.03	1M	131K	fp4	97.4%
Mistralcache mistral	$0.15	$0.50	$0.03	1M	131K	—	100.0%
KTcache kt	$0.15	$0.50	$0.03	1M	944K	—	100.0%
Coherecache cohere/nvfp4	$0.15	$0.50	$0.03	1M	131K	nvfp4	99.6%
Nebiuscache nebius/fp4	$0.15	$0.50	$0.03	1M	944K	fp4	99.7%
Modalcache modal/fp4	$0.15	$0.50	$0.03	1M	944K	fp4	100.0%
BaseTencache baseten/fp8/us	$0.15	$0.50	$0.03	1M	131K	fp8	0.0%
Crusoecache crusoe/fp4	$0.15	$0.50	$0.03	1M	944K	fp4	100.0%
AtlasCloudcache atlas-cloud/fp8	$0.15	$0.50	$0.03	1M	131K	fp8	100.0%
Fireworkscache fireworks	$0.15	$0.50	$0.03	1M	944K	—	99.8%
Friendlicache friendli	$0.15	$0.50	$0.03	1M	944K	—	100.0%
SiliconFlowcache siliconflow/fp8	$0.15	$0.50	$0.03	1M	262K	fp8	99.3%
Togethercache together	$0.15	$0.50	$0.03	1M	944K	—	100.0%
Parasailcache parasail/fp4	$0.15	$0.50	$0.03	1M	944K	fp4	100.0%
BaseTencache baseten/fp8	$0.15	$0.50	$0.03	1M	131K	fp8	98.5%
Venicecache venice	$0.15	$0.50	$0.03	1M	131K	—	99.6%
DeepInfracache deepinfra/fp4	$0.15	$0.50	$0.03	1M	944K	fp4	100.0%
Z.AIcache z-ai/fp8	$0.15	$0.50	$0.03	1M	131K	fp8	100.0%
Inceptroncache inceptron/fp8	$0.175	$0.50	$0.03	1M	944K	fp8	100.0%
Sail Researchcache sail-research/us	$0.175	$0.625	$0.022	1M	131K	fp4	100.0%
Sail Researchcache sail-research/fp4	$0.175	$0.625	$0.022	1M	131K	fp4	100.0%
Parasailcache parasail/fast	$0.188	$0.625	$0.037	1M	944K	fp4	100.0%

Supported API Parameters

Parameters accepted by this model when called through ChatOSS or OpenRouter.

frequency_penaltyinclude_reasoninglogit_biaslogprobsmax_tokensmin_pparallel_tool_callspresence_penaltyreasoningreasoning_effortrepetition_penaltyresponse_formatseedstopstructured_outputstemperaturetool_choicetoolstop_ktop_logprobstop_p

Deep-Dive Benchmarks & Teardowns

Independent benchmark analysis and real-world agent billing teardowns featuring GLM 5.3 Flash.

2026-09-21 · 11 min read

GLM 5.3 Flash vs. DeepSeek V4.1 Flash: which is better at agentic coding?

Both labs ship a million-token context, MIT weights, and a ~$0.15-per-million price tag. On the independent harnesses that have actually run both models — vals.ai's Vibe Code Bench and Terminal-Bench 2.1, BenchLM's Vibe Code Bench 1-100, and Artificial Analysis' v4.3 index — DeepSeek V4.1 Flash wins agentic coding, and by a wider margin than either vendor's own table suggests.

Read article →

Compare Similar Models

Browse all models →

Z.ai (GLM)

GLM 5.3 FlashX Legal

GLM-5.3-FlashX-Legal (GLM-Legal) is Z.ai's legal-domain variant of GLM-5.3-FlashX, a native multimodal model that accepts text, image, and video input and produces text. It adds legal-domain mid-training and post-training with reinforcement...

$0.37 / $1.25

Z.ai (GLM)

GLM 5.3 Prime

GLM-5.3-Prime is the high-speed variant of Z.ai's GLM-5.3, inheriting its full capabilities while delivering 1.5–2× the output throughput through inference acceleration. It supports text input and output with a 1M-token...

$2.80 / $8.80Coding 58.3%

Z.ai (GLM)

GLM 5.3 FlashX

GLM-5.3-FlashX is the high-speed variant of Z.ai's GLM-5.3-Flash, a native multimodal model delivering inference speeds of up to 200 tokens/s. Built on the same hybrid sparse and linear attention architecture...

$0.37 / $1.25Coding 36.9%

Z.ai (GLM)

GLM 5.3

GLM-5.3 is a large-scale reasoning model from Z.ai, built for complex software engineering and long-horizon agent tasks. It supports text input and output with a 1M-token context window, and improves...

$0.07 / $1.40Coding 58.3%

Anthropic

Claude Opus 5.5

Claude Opus 5.5 is Anthropic's flagship model for demanding reasoning, coding, and long-horizon agentic work, succeeding Claude Opus 5. It is particularly strong at multi-step changes in large codebases, code...

$4.00 / $20.00Coding 74.1%

Moonshot AI

Kimi K3.1 Pro

Kimi K3.1 Pro is Moonshot AI's flagship open-weight multimodal reasoning model and a post-trained successor to Kimi K3, with a 1M-token context window. It is built for long-horizon coding, 3D...

$2.00 / $10.00Coding 65%