GLM 5.1
glm-5-1glm-5.1z-ai/glm-5.1zai-org-glm-5-1- π§ Next-generation flagship built for agentic engineering and long-horizon tasks.
- π§ Designed to keep using tools across extended autonomous runs.
- π’ Built by Z.ai (formerly Zhipu AI), released under MIT license.
- π Roughly 200K-token context window for extended documents and repos.
- π Incremental refresh of GLM 5 with stronger coding and reasoning.
- β‘ Ships as an FP8 checkpoint for efficient serving.
- π¬ Supports reasoning traces, function calling, and web search.
- π Open weights published on Hugging Face under MIT.
Z.ai, formally Knowledge Atlas Technology Joint Stock Co., Ltd., is a Chinese technology company specializing in artificial intelligence. Previously known internationally as Zhipu AI, the company rebranded to Z.ai in 2025. Its core focus is the GLM family of large languageβ¦
Explore 17 more models by Z.ai βGLM 5.1 is Z.ai's next-generation flagship large language model, positioned for agentic engineering, complex software tasks, and long-horizon planning with extended tool use. It is released under the MIT license, continuing the open-weight strategy the company (formerly Zhipu AI) has applied across its GLM family. The catalog lists a roughly 200K-token context window, and the model ships in an FP8 quantization for efficient deployment.
Z.ai describes GLM 5.1 as an evolution of GLM 5, built to keep revising and improving across long autonomous runs rather than plateauing early. In its release notes, the company states that, relative to GLM 5, version 5.1 delivers gains in coding, agentic tool use, reasoning, and long-horizon agentic tasks. It is further tuned for agentic coding workflows, with adjustments to its chat template for deferred tool loading.
On Z.ai's own reported benchmarks, GLM 5.1 scores 58.4 on SWE-Bench Pro, and the company says it improves over its predecessor GLM 5 across major coding, mathematical reasoning, and agentic evaluations. Capabilities exposed through the catalog include reasoning, function calling, and web search.
In the family timeline, GLM 5.1 follows GLM 5 and earlier releases such as GLM 4.7, and precedes the later GLM 5.2, which Z.ai positions as a further step in long-horizon capability.
This About section is AI-generated from public sources via VeniceStats + Venice inference, with no human editing. It may contain inaccuracies.
Daily usageΒ Β·Β GLM 5.1Β Β·Β Organic providers
Top providersΒ Β·Β GLM 5.1Β Β·Β Organic
- Open Forge47.8%
- Phala26.7%
- π¬ Banter5.1%
- Super Seeder4.9%
- Best3.5%
- Relay Works2.0%
- Others10.0%
| Provider | Health | Repβ | Sales | Tokens | Routing | Input $/M | Cached $/M | Output $/M | Save | 1st token | tok/s |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Phala 0x88c8β¦15d6 | healthy | 95.06 | $146.35 | 87.00M | #3 | $1.21 | $0.60 | $4.20 | β | β | β |
| Super Seeder 0xd19fβ¦41f3 | healthy | 90.57 | $26.81 | 33.01M | #5 | $1.4517 | $0.2696 | $4.5626 | β | β | β |
| β² Apex Ant 0x73b4β¦e736 | healthy | 88.20 | $3.50 | 10.34M | #1 | $0.77 | $0.14 | $2.42 | β35% | β | β |
| Venice.ai Proxy 0x1f22β¦18c9 | healthy | 75.07 | $9.93 | 9.39M | #2 | $0.847 | $0.1573 | $2.662 | β29% | β | β |
| RedPill 0x7e99β¦dc4f | healthy | 65.36 | $6.87 | 4.19M | #4 | $1.21 | $0.60 | $4.20 | β | β | β |
| Relay Works 0x3040β¦7139 | β | 58.03 | $11.24 | 11.28M | gated | $0.585 | $0.077 | $1.885 | β51% | β | β |
| HallPass 0xcac8β¦71b4 | β | 20.00 | $0.62 | 277.66k | gated | $0.03 | $0.006 | $0.095 | β97% | β | β |
| klinn 0x0ee1β¦6c41 | β | 12.34 | $1.65 | 9.51M | gated | $0.0294 | $0.0059 | $0.0931 | β98% | β | β |
| Chutes 0xded6β¦657c | β | 7.85 | $0.27 | 129.74k | gated | $1.078 | $0.1078 | $3.388 | β9% | β | β |
| dealer 0xc46eβ¦8ce6 | β | 2.11 | $0.60 | 269.78k | gated | $1.33 | $0.247 | $4.18 | β | β | β |
| ChainScout AI 0x1734β¦e621 | β | 0.00 | β | β | gated | $2.00 | $2.00not advertised | $6.00 | β | β | β |
| surplus-hive 0x6145β¦5c58 | β | 0.00 | $0.12 | 26.22k | gated | $1.8292 | $1.8292 | $5.7491 | β | β | β |
| Fire Ant π₯π 0xbe05β¦bc5d | β | 0.00 | $8.75 | 6.33M | gated | $1.386 | $0.2009 | $3.49 | β | β | β |
| uomi.ai 0x87dfβ¦48e3 | β | 0.00 | β | β | gated | $1.379 | $1.379not advertised | $4.374 | β | β | β |
| minion0x 0x215eβ¦e2e3 | β | 0.00 | β | β | gated | $1.03 | $0.20 | $3.48 | β13% | β | β |
| Best 0x472fβ¦69fd | healthy | 0.00 | $19.11 | 49.25M | gated | $1.0006 | $0.8505 | $3.3575 | β16% | 0.2 s | 210 |
| Open Bird 0xc0f1β¦8183 | β | 0.00 | $11.18 | 8.74M | gated | $0.77 | $0.143 | $2.42 | β35% | β | β |
| antseed-neon-puma-944e 0x6650β¦944e | β | 0.00 | β | β | gated | $0.5968 | $0.325 | $1.8755 | β50% | β | β |
| Joi Gateway 0x8a90β¦7007 | β | 0.00 | β | β | gated | $0.5893 | $0.5893 | $1.852 | β50% | β | β |
| LLMart 0x948aβ¦a000 | β | 0.00 | $2.76 | 5.81M | gated | $0.5532 | $0.1479 | $1.9168 | β54% | β | β |
| Leftermute 0x388bβ¦5389 | β | 0.00 | β | β | gated | $0.2121 | $0.2121not advertised | $0.6666 | β82% | β | β |
| Apex TEE Test 0xe672β¦7955 | β | 0.00 | β | β | gated | $0.1944 | $0.189 | $0.648 | β84% | β | β |
| Meridian AI 0x8c8cβ¦06f5 | β | 0.00 | $1.05 | 3.24M | gated | $0.1497 | $0.1497not advertised | $0.4704 | β87% | β | β |
| surplus-hive 0xe177β¦e91c | β | 0.00 | $0.05 | 439.29k | gated | $0.0532 | $0.0532 | $0.1672 | β96% | β | β |
| NovaRoute AI 0xc50dβ¦ed7b | β | 0.00 | $1.35 | 1.87M | gated | $0.0365 | $0.3812 | $0.1147 | β97% | β | β |
"Best price" and the provider table are live Antseed catalog data (advertised $/1M tokens β or $ per generated image for unit-billed image models β not settled amounts). Reputation = buyer trust score (0-100, the Antseed SDK's own formula). "Routing" = the SDK's default buyer routing (what Antseed Desktop ships with): a trust β₯ 60 gate on the effective reputation, then cheapest-first among routable providers; live failover state (per-peer cooldowns) is buyer-side runtime and not included. Model knowledge (TLDR, provider, About) via the VeniceStats enrichment layer. Advertised catalog, not the model used in any specific purchase. "Usage on Antseed" counts only settlements whose buyers share the per-model split on-chain (metadata v2/v3, opt-in), so every usage figure is a lower bound.