- Type
- Open weightsMIT License
- Params
- 754B
- Context
- 205K
about 154K words of context
Our take
Written Aug 3, 2026GLM 5.1 is a large downloadable text model from Z.AI with a permissive MIT licence and a 204,800-token request limit. Its strongest measured skill is coding, though creative writing lags well behind, and no active-parameter count is disclosed.
Pick this when you need a 754-billion-parameter model with a genuinely permissive licence for commercial use, modification or redistribution. Use it for coding workloads where its measured Arena score peaks, or for general text tasks within its 204,800-token limit. Skip it if you need image, audio or video input, if creative writing quality is critical, or if you want to optimise costs aggressively — the price spread between hosts is narrow.
The case for it
- Strongest measured skill is coding, with an Arena Coding Elo 49.6 points above its own overall text score.
- Permissive MIT licence allows commercial use, modification and redistribution with minimal attribution.
- Ten offers from nine providers, with six at sub-dollar input rates.
- A high-throughput option exists: one host delivers 81 tokens per second, 2.7 times the speed of the cheapest alternative.
The case against it
- No disclosed active-parameter count or efficiency architecture; 753.9 billion total parameters with no information on how many activate per token.
- Creative writing is its weakest measured skill, 70.3 points below its coding score on the same Arena scale.
- Throughput data is incomplete: two of ten offers lack speed figures.
How good is it?
IntelligencePuzzles, maths, exam questions
Arena Text (overall)21st of 143 · 1468.8
CodingWriting and fixing code on its own
Arena Coding16th of 143 · 1518.4
AgenticPlanning, calling tools, staying on task
Arena Agent (IPS)17th of 36 · 0.004
Arena Agent (IPS) is the only board that has scored it for this.
WritingWe do not rate this
Two boards come close and neither tests writing: Arena Creative Writing asks people which of two replies they prefer, and LiveBench Language tests whether a model understood a passage. So we show where GLM 5.1 placed and give it no mark out of five.
These tests check whether a model follows instructions — a precondition for all the work above, but not a measure of how well that work is done, which is why they get no rating.
Every published score for this model8 scoresEvery figure we hold, from 8 boards, with who ran it and a link to the source — including the boards no rating above is built on.
Can you run it yourself?
- Fits in memory
- weights load entirely on the card
- Spills to system RAM
- some weights offload; much slower
- Too large
- will not load even with offload
- est
- size is calculated; the verdict could change by 10%
GeForce RTX 4090 · 24 GB
Too large for this card. The weights do not fit even with part of them offloaded to system memory.
Apple M1 Pro (16-core GPU) · 32 GB
Too large for this card. The weights do not fit even with part of them offloaded to system memory.
Radeon RX 7900 XT · 20 GB
Too large for this card. The weights do not fit even with part of them offloaded to system memory.
Memory use by level
Against a 24 GB card.
All 3 sizes here are calculated, not measured. We hold no measured file for this model, so each size comes from the parameter count and every verdict above inherits that uncertainty.
Check against your own machine → · All 71 devices, with every size →
Or rent it from someone else
Cheapest of 23 live listings. Picked at the widest standard context we hold, within one quantisation slice, so the numbers beside it are a price one host actually charges.
- per 1M tokens
- $0.97 in / $3.04 out
- Context served
- 205K
- Throughput
- Not measured
| Provider | In / out per 1M tokens | Context | Throughput | Trains on prompts | Logs prompts | Zero retention |
|---|---|---|---|---|---|---|
| Baidufp8 | $0.95 / $2.99 | 203K | 30 tok/s | No | Yesunknown period | Unknown |
| StreamLakefp8 | $0.97 / $3.04 | 200K | 56 tok/s | No | Yesunknown period | Unknown |
| OpenRouter | $0.97 / $3.04 | 205K | not measured | Unknown | Unknown | Unknown |
| GMICloudfp8 | $0.98 / $3.08 | 203K | 44 tok/s | No | Yesunknown period | Unknown |
| Chutesfp8 | $0.98 / $3.08 | 203K | 27 tok/s | No | Yesunknown period | Unknown |
| Waferfp4 | $1.00 / $3.20 | 203K | 81 tok/s | No | No | Confirmed |
| DeepInfrafp4 | $1.05 / $3.50 | 203K | 31 tok/s | No | No | Confirmed |
| DeepInfrafp4 | $1.05 / $3.50 | 203K | not measured | Unknown | Unknown | Unknown |
| SiliconFlowfp8 | $1.19 / $3.74 | 205K | 53 tok/s | No | No | Confirmed |
| AtlasCloudfp8 | $1.26 / $3.96 | 203K | 38 tok/s | No | Yesunknown period | Unknown |
| Alibaba Cloudfp8 | $1.33 / $4.18 | 203K | 51 tok/s | No | Yesunknown period | Unknown |
| Phala | $1.21 / $4.20 | 203K | 27 tok/s | No | No | Confirmed |
| DigitalOcean Gradient | $0.97 / $4.30 | 164K | 18 tok/s | No | No | Confirmed |
| Z.AIfp8 | $1.40 / $4.40 | 203K | 52 tok/s | No | No | Confirmed |
| Novita AI | $1.38 / $4.40 | 205K | not measured | Unknown | Unknown | Unknown |
| Novita AIfp8 | $1.38 / $4.40 | 205K | 37 tok/s | No | No | Confirmed |
| Crusoefp8 | $1.20 / $4.40 | 203K | 73 tok/s | No | No | Confirmed |
| Parasailfp8 | $1.40 / $4.40 | 203K | 68 tok/s | No | No | Confirmed |
| CoreWeavefp8 | $1.40 / $4.40 | 203K | 72 tok/s | No | No | Confirmed |
| Nebius AI Studiofp8 | $1.40 / $4.40 | 203K | 33 tok/s | No | No | Confirmed |
| Fireworks AI | $1.40 / $4.40 | 203K | not measured | No | No | Unknown |
| Friendli | $1.40 / $4.40 | 203K | 71 tok/s | No | Yesunknown period | Unknown |
| Venice AIfp8 | $1.54 / $4.84 | 200K | 26 tok/s | No | No | Confirmed |
Baidu
DeepInfra
DigitalOcean Gradient
Novita AI
Nebius AI Studio
Fireworks AIAcross the 23 listings we hold: 20 say they do not train on prompts, 0 say they do and 3 do not say. 12 appear in the zero-retention registry we check; the rest are unknown to us rather than confirmed either way.
What each host's API supports
From the parameter list each endpoint publishes. Streaming is omitted: nothing we hold reports it, for any model.
- ✓
- Supported
- ✗
- Not supported
- Not published
- host gave no parameter list
| Provider | Tool calling | JSON output | Strict schema |
|---|---|---|---|
| Baidufp8 | ✓ | ✓ | ✓ |
| StreamLakefp8 | ✓ | ✓ | ✓ |
| OpenRouter | ✓ | ✓ | ✓ |
| GMICloudfp8 | ✗ | ✗ | ✗ |
| Chutesfp8 | ✓ | ✓ | ✓ |
| Waferfp4 | ✓ | ✓ | ✓ |
| DeepInfrafp4 | ✓ | ✓ | ✓ |
| DeepInfrafp4 | |||
| SiliconFlowfp8 | ✓ | ✗ | ✗ |
| AtlasCloudfp8 | ✓ | ✓ | ✓ |
| Alibaba Cloudfp8 | ✓ | ✓ | ✓ |
| Phala | ✓ | ✓ | ✓ |
| DigitalOcean Gradient | ✓ | ✓ | ✓ |
| Z.AIfp8 | ✓ | ✓ | ✗ |
| Novita AI | |||
| Novita AIfp8 | ✓ | ✗ | ✗ |
| Crusoefp8 | ✓ | ✓ | ✗ |
| Parasailfp8 | ✓ | ✓ | ✓ |
| CoreWeavefp8 | ✓ | ✓ | ✗ |
| Nebius AI Studiofp8 | ✓ | ✓ | ✓ |
| Fireworks AI | ✓ | ✓ | ✓ |
| Friendli | ✓ | ✓ | ✓ |
| Venice AIfp8 | ✓ | ✓ | ✓ |
Tool calling: 20 of 23 listings say yes, 1 says no, 2 publish no parameter list. JSON output: 18 of 23 listings say yes, 3 say no, 2 publish no parameter list. Strict schema: 15 of 23 listings say yes, 6 say no, 2 publish no parameter list.
Models people weigh against GLM 5.1
When we formed this view
Dates behind this page
Prices last checked 32h ago
What we do not know about this model yet
- We hold no measured file for it, so all 3 sizes on this page are calculated from the parameter count.
- 2 of 23 listings publish no parameter list, so what their API accepts is unknown to us.
- Nothing we hold says whether an endpoint streams, so we do not show it either way.
- 3 of 23 listings do not say whether they train on prompts.
Licence and identifiers
What the licence allowsMIT License, what it allows commercially, and the identifiers you need to pull this model — its Hugging Face repo, our slug and a machine-readable card.
Licence
MIT License
Fully permissive: do anything with attribution. No patent grant, unlike Apache-2.0.
Identifiers
- Hugging Face
- zai-org/GLM-5.1
- Architecture
- Mixture of experts
- Modality record
- text->text
- Catalogue slug
- z-ai-glm-5-1