Z.ai · in Borg
Z.aiGLM-5.2
GLM-5.2 is the long-context successor to GLM-5.1, retaining a focus on sustained engineering but substantially expanding how much project material one request can hold.
Reasoning1,048,576 tokenstext
Included in Borg
Download Borg Borg for BusinessAt a glance
| Specification | Value |
|---|---|
| Provider | Z.ai |
| Family | GLM |
| Release date | 2026-06-16 Vendor source |
| Context window | 1,048,576 tokens Vendor source |
| Input modalities | text |
| Output modalities | text |
| Reasoning | Yes |
| Open weights | Yes |
| Borg lane ids | glm-5.2, z-ai/glm-5.2 |
| Versions in Borg | GLM-4.5 · GLM-4.6 · GLM-4.7 · GLM-5 · GLM-5.1 · GLM-5.2 · GLM-5.3 |
| Median time to first token (gateway proxy) on Borg | 7068 ms |
| Median tokens per second (end to end) on Borg | 25.7 output tokens per second |
Not published by the vendor: thinking modes, knowledge cutoff, fusion roles.
What GLM-5.2 is good for
GLM-5.2 is the long-context successor to GLM-5.1, retaining a focus on sustained engineering but substantially expanding how much project material one request can hold. Z.ai describes whole-codebase understanding, preservation of module boundaries and architectural constraints, and end-to-end refactoring as primary uses. It is a text-in/text-out flagship with thinking and tool calling, not a vision endpoint. Its bigger window distinguishes it from GLM-5 and GLM-5.1; GLM-5.3 later reuses its base model while improving complex coding through post-training. This model is well suited to repository audits, cross-file migrations and agents that need to keep a technical plan coherent over long sessions. Borg includes it in the subscription; Fusion can route suitable work to it, or you can pick its lane directly.
- whole-repository architecture audits
- cross-file refactoring
- long-context engineering agents
- requirements-to-deployment planning
How Borg uses it
GLM-5.2 is available as a subscription catalog lane in Borg. The catalog does not publish which Fusion profiles route to this model, so no routing role is claimed here. You can pick it directly by setting the OpenAI-compatible gateway model field to a listed lane id. Check the current catalog for availability before building a workflow around a lane.
OpenAImodel: "glm-5.2"Measured on Borg
| Metric | Median | 95th percentile | Samples |
|---|---|---|---|
| Time to first token (gateway proxy) | 7068 ms | 22326 ms | 32197 |
| Tokens per second (end to end) | 25.7 output tokens per second | 72.9 output tokens per second | 304255 |
Time to first token (gateway proxy)
Output tokens per second (whole turn)
- Time to first token (gateway proxy): time until the first streamed response event reaches the gateway; our closest available measure of time to first token
- Tokens per second (end to end): tokens per second, end to end: provider-reported output tokens divided by the whole turn's elapsed time, including queueing and reasoning; not a streaming per-token rate
Completion speed covers the whole turn including queueing and reasoning time; it is not a streaming tokens-per-second figure.
Measured on Borg: SearchAtlas AI Gateway production request logs. 30-day window ending 2026-10-08; 333985 requests. Minimum 100 samples per published request metric.
Benchmarks
| Benchmark | Vendor-reported score | Source |
|---|---|---|
| Terminal-Bench 2.1 | 81 | Vendor source |
| SWE-bench Pro | 62.1 | Vendor source |
| AIME 2026 | 99.2 | Vendor source |
| GPQA-Diamond | 91.2 | Vendor source |
Borg coding bench
Hard-graded with tests plus lint, no LLM judge; 3 tasks, 15 runs, 2026-09-20.
| Metric | Pass rate |
|---|---|
| Overall pass rate | 47% |
| Hidden-test pass rate | 53% |
Source: Borg coding bench, hard-graded by tests and ruff; no LLM judge
Pricing
Included in Borg subscription. Vendor API list price: $1.4 input / $4.4 output per 1M tokens. Vendor price source.
Compare subscription and API pricing · Cost Savings Calculator
FAQ
What is GLM-5.2?
GLM-5.2 is listed in Borg's subscription catalog as glm-5.2. GLM-5.2 is the long-context successor to GLM-5.1, retaining a focus on sustained engineering but substantially expanding how much project material one request can hold.
Is GLM-5.2 in Borg?
Yes. GLM-5.2 is listed in Borg's subscription catalog under the lane glm-5.2. You can choose the lane directly, or explore the available Fusion profiles. Availability and model details can change; check the current catalog before deployment.
What context window does GLM-5.2 have?
The published context window is 1,048,576 tokens. Consult the cited sources for details and current limits. Vendor source
Is GLM-5.2 good for coding?
Vendor-published evaluations appear in the benchmarks section; their settings may differ from your tasks. Its listed uses include whole-repository architecture audits, cross-file refactoring, long-context engineering agents, requirements-to-deployment planning. Try the lane for your own tasks, and compare the output with your requirements before choosing a default.
How fast is GLM-5.2 on Borg?
Measured on Borg: SearchAtlas AI Gateway production request logs. 30-day window ending 2026-10-08; 333985 requests. Minimum 100 samples per published request metric.
How does GLM-5.2 do on Borg's coding bench?
GLM-5.2 passed 47% overall and 53% of hidden tests across 15 runs.
How much does GLM-5.2 cost on Borg vs the API?
This model is included in the Borg subscription. The vendor API list prices appear below with a source link. See the pricing explainer and cost savings calculator to compare billing approaches.
Sources
Related
Other versions of GLM: GLM-4.5 · GLM-4.6 · GLM-4.7 · GLM-5 · GLM-5.1 · GLM-5.3
Other Z.ai models: GLM-4.5 · GLM-4.5-Air · GLM-4.6 · GLM-4.7 · GLM-5 · GLM-5-Turbo · GLM-5.1 · GLM-5.3
Models sharing a Fusion role: Not published