Loading page
Zhipu
GLM-5.1
GLM-5.1 is a large language model from Zhipu (Z.ai) in the GLM-5 family, offering a 200,000-token context window.
Context 200K
Input from
$0.966 / 1M tokens
across 8 providers
API Pricing
Cheapest on OpenRouter — 20% below avg| Provider | Input / 1M | Output / 1M | Cached / 1M |
|---|---|---|---|
| $0.966 | $3.04 | $0.179 | |
| $1.00 | $3.20 | - | |
| $1.05 | $3.50 | $0.205 | |
| $1.22 | $3.96 | $0.610 | |
| $1.30 | $4.30 | $0.260 | |
| $1.38 | $4.40 | $0.260 | |
| $1.40 | $4.40 | - | |
| $1.40 | $4.40 | - |
Prices updated daily. Last check: Sep 6, 2026
GLM-5.1 pricing by provider
Input, output, and batch rates, plus alternatives, for one provider at a time.
Performance & Benchmarks
Source: Artificial Analysis →Intelligence
28.5 / 100
Reasoning & Knowledge
- GPQA Diamond83.9%
- Humanity's Last Exam27.9%
Agentic & Tool Use
- Terminal-Bench Hard35.6%
- τ²-bench97.1%
Instruction & Long Context
- IFBench52.0%
- Long-Context Reasoning53.3%
Benchmarks measured Sep 2026. Scores are independent evaluations, not vendor-reported.
Model Details
General
- Creator
- Zhipu
- Family
- GLM-5
- Context Window
- 200K
- Modalities
- Text
Capabilities
- Tool Calling
- No
- Open Source
- No
- Aliases
- glm-5.1, GLM-5.1, zai-org/GLM-5.1, THUDM/GLM-5.1
Strengths & Limitations
Strengths
- 200,000-token context window supports long documents, large diffs, and extended conversation histories in a single request
- Part of Zhipu's GLM-5 family, an incremental revision over the initial GLM-5 release
- Published under multiple aliases (zai-org/GLM-5.1, THUDM/GLM-5.1), making it identifiable across different inference catalogs
- Available through more than one inference endpoint, allowing price and latency comparison between hosts
- GLM series models are commonly used for both Chinese and English workloads, which suits bilingual applications
- Large context reduces reliance on complex chunking or retrieval pipelines for long-input tasks
Limitations
- We do not have confirmed throughput or time-to-first-token measurements for GLM-5.1, so serving speed must be validated per provider
- Context window of 200K tokens is smaller than the 1M-token windows some competing long-context models advertise
- Modality support and tool-calling details are not tracked in our dataset and vary by host
- Provider coverage for GLM models is narrower than for the largest Western model families, which can limit regional availability and failover options
- Benchmark results in our dataset are incomplete, making direct quality comparison with peers harder
Key Features
•200,000-token context window
•GLM-5 family model from Zhipu (Z.ai)
•Multiple published aliases: zai-org/GLM-5.1 and THUDM/GLM-5.1
•Available via multiple third-party inference providers
•Long-context document and codebase analysis
•Chinese and English language usage, consistent with the GLM series
•API access through Zhipu's own platform and hosted endpoints
About GLM-5.1
GLM-5.1 is a large language model released by Zhipu (also published under the zai-org and THUDM organization names) as part of the GLM-5 family. It sits as an incremental revision within that family rather than a separate line, and is distributed under identifiers such as zai-org/GLM-5.1 and THUDM/GLM-5.1, which are the names typically used by inference providers that host it.
The main confirmed specification we track for GLM-5.1 is its 200,000-token context window, which is large enough to hold long documents, extended chat histories, or sizeable portions of a codebase in a single request. Zhipu's GLM series has historically been associated with bilingual Chinese and English usage, and GLM-5.1 is served through multiple third-party inference endpoints in addition to Zhipu's own API. Other attributes — throughput, latency, modality support, and tool-calling specifics — are not fully captured in our dataset, so they are best verified against the provider you plan to use.
In practice, models in this class are used for long-context document work, coding assistance, and general assistant workloads where a wide context window reduces the need for aggressive chunking or retrieval engineering. Because GLM-5.1 is offered by several hosts, throughput and pricing can differ substantially between providers even for the same model weights; the pricing table on this page tracks the options we have collected.
Common Use Cases
GLM-5.1 fits workloads that benefit from a wide context window: analyzing long contracts, reports, or research papers in one pass; reviewing multi-file code changes; summarizing long transcripts; and powering assistants that need to retain lengthy conversation history. Its position in Zhipu's GLM-5 family and the series' bilingual heritage make it a candidate for applications serving Chinese and English users, or for teams evaluating non-US model providers for cost or sourcing reasons. Because it is hosted by several inference providers, it also suits teams that want the option to move the same model between endpoints. For latency-sensitive or extremely high-volume classification work, benchmark it against smaller models first, since we do not track confirmed throughput figures for GLM-5.1.
Frequently Asked Questions
How much does GLM-5.1 cost to use?
Pricing depends on which provider hosts the model and on the pricing type — input versus output tokens, cached input, and any batch or committed-use discounts. Rates for the same model can differ meaningfully between endpoints. See the pricing table on this page for the current per-provider figures we track.
What is GLM-5.1 best used for?
It suits long-context tasks such as document analysis, multi-file code review, long transcript summarization, and assistants that carry extended conversation history, thanks to its 200,000-token context window. Its GLM lineage also makes it a common choice for bilingual Chinese and English applications.
How large is the context window?
GLM-5.1 supports a 200,000-token context window. That is enough for several hundred pages of text or a substantial portion of a codebase in a single request, though it is smaller than the 1M-token windows some competing long-context models offer.
How does GLM-5.1 relate to GLM-5?
GLM-5.1 is part of the same GLM-5 family from Zhipu and represents an incremental revision within that line rather than a separate model series. If you are choosing between them, test both on your own prompts, since our dataset does not include complete comparative benchmarks.
Why does GLM-5.1 appear under different names?
Providers list the model under several identifiers, including glm-5.1, zai-org/GLM-5.1, and THUDM/GLM-5.1. These refer to the same model; THUDM and zai-org are organization names associated with Zhipu's published releases.