IntakeAnyone files from Claude or ChatGPT, in a real conversation.The board and DocsEvery request with its brief, size, worth and seat, and the specs and decisions beside it.Smart routingThe best seat and the best model for every request.The recordEstimate against actual, signed. Better with every row.ConnectionsClaude, ChatGPT, Claude Code, Codex, GitHub, Slack and more.
Business teamsSales, success, ops: send the customer's request and hear when it lands.Product teamsSpecs and prototypes that arrive ready to build, and a say in what runs.EngineeringThe right work, on the right seat and model. For engineers and leads.
Docs and helpEvery verb, every surface.Developerscurl -fsSL sleds.ai/install | shTrustWhat we keep, and the exceptions.ManifestoWhat we are doing, and why.Claude connectormcp.sleds.ai/mcpSelf-hostThe record in your cloud, on your keys.ContactTalk to the people building it.
Request an inviteSign in
sleds/Model Standings/GLM-5.3
Zhipu · open weights

GLM-5.3.

GLM-5.3 by Zhipu: P8 on the Nov 3, 2026 board with a score of 45, 13 behind Opus 5.5. List price $1.40 in and $4.40 out per million tokens.

timing by
sleds.ai
as of Nov 3, 2026the full board →
PositionP8on the latest board
Score45Artificial Analysis Intelligence Index
Price$1.40 / $4.40per 1M tokens in / out · $2.15 blended
Test lapnot yeta verdict from real work
score over time
24 daysunchanged
blended price over time
24 daysunchanged
history · GLM-5.3newest first
DayPosScore$ in / outOpenRouterVercelSpendTest lap
Nov 3, 2026P845$1.40 / $4.401.8%6.9%9.2%–
Nov 2, 2026P845$1.40 / $4.401.9%7.2%9.7%–
Nov 1, 2026P845$1.40 / $4.402.2%8.0%11%–
Oct 31, 2026P845$1.40 / $4.402.3%9.1%12%–
Oct 30, 2026P845$1.40 / $4.402.4%10.0%13%–
Oct 29, 2026P845$1.40 / $4.402.6%10%14%–
Oct 28, 2026P845$1.40 / $4.402.8%11%15%–
Oct 27, 2026P845$1.40 / $4.402.7%11%16%–
Oct 26, 2026P845$1.40 / $4.402.7%12%17%–
Oct 25, 2026P845$1.40 / $4.402.7%12%17%–
Oct 24, 2026P845$1.40 / $4.402.7%11%17%–
Oct 23, 2026P845$1.40 / $4.402.8%10%17%–
Oct 22, 2026P845$1.40 / $4.403.0%11%18%–
Oct 21, 2026P845$1.40 / $4.403.1%11%20%–
Oct 20, 2026P845$1.40 / $4.403.3%11%20%–
Oct 19, 2026P845$1.40 / $4.403.3%11%21%–
Oct 18, 2026P845$1.40 / $4.403.2%11%21%–
Oct 17, 2026P845$1.40 / $4.403.2%11%19%–
Oct 16, 2026P845$1.40 / $4.403.0%9.8%17%–
Oct 15, 2026P845$1.40 / $4.402.7%8.4%14%–
Oct 14, 2026P845$1.40 / $4.402.5%7.3%12%–
Oct 13, 2026P845$1.40 / $4.40–6.2%9.2%–
Oct 12, 2026P845$1.40 / $4.40–4.8%6.5%–
Oct 11, 2026P845$1.40 / $4.40–3.9%5.1%–
the spec sheet
Context window
1M tokens
Max output
1M tokens
Released
Aug 18, 2026
Inputs
text
Outputs
text
Cached input
$0.14 per 1M tokens
Weights
open
questions people ask
How much does GLM-5.3 cost?

$1.40 per million input tokens and $4.40 per million output tokens at list price, $2.15 blended at three input tokens to one output. Cached input is $0.14 per million.

How does GLM-5.3 score?

45 on the Artificial Analysis Intelligence Index: P8 of 14 on the Nov 3, 2026 board, 13 behind Opus 5.5.

What is the context window of GLM-5.3?

1M tokens, with up to 1M tokens of output.

How widely is GLM-5.3 used?

Over the trailing seven days: 1.8% of tokens on OpenRouter; 6.9% of tokens on Vercel AI Gateway; 9.2% of spend on Vercel AI Gateway.

Is GLM-5.3 good for real work?

It has not had a Test Lap yet, our verdict from putting a model on real work for a while.

the rest of the grid
Opus 5.5Sonnet 5.5Gemini 4 ArgonFable 5.1GPT-6 AstraGPT-6.1 SolGrok 4.7Kimi K3Haiku 5.5GPT-6 LunaV4.1 FlashGemini 3.8 FlashMuse Spark 1.3

Score is the Artificial Analysis Intelligence Index. Price is list, per million tokens. Usage is the share of tokens over the trailing seven days on OpenRouter and on the Vercel AI Gateway usage data, both licensed under CC BY 4.0; a dash means too small to call. Specs come from Vercel AI Gateway's model list and the lab's own documentation. Timing by sleds, which routes work across every lab and favours none.