Coding Models
Models tuned for code completion, refactoring, and review. Stronger on programming benchmarks than general LLMs of similar parameter count, while remaining usable through the standard Chat Completions endpoint.
Showing 4 of 4 models
| Token-by-token output via Server-Sent Events. Suitable for low-latency, real-time UI. | Function / tool calling (OpenAI-compatible). The model can return structured tool invocations. | Accepted values of the reasoning_effort parameter, from the platform model datasheet. "-": no values are published for this model. See the Reasoning Control guide. | Where the response reports the number of reasoning tokens in the usage object: completion_tokens_details.reasoning_tokens, usage.reasoning_tokens, or not reported (the model reasons, but the response carries no count). "-": the model has no reasoning. | Price per million prompt tokens served from the prefix cache: 15% of the normal input price. T-Cloud-hosted models only — cached pricing for external providers will be added later. | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| GLM 5.2 |  Zhipu AI | Hosted on T-Cloud Public — Telekom's sovereign infrastructure in Germany. | Text | Text | ✓ | ✓ | 1M | nonehighmax | usage | €1.50 | €3.50 | €0.225 | EssentialProfessionalAgentic | |
| GLM 5.3 Flash | Zhipu AI | Hosted on T-Cloud Public — Telekom's sovereign infrastructure in Germany. | Text, Image | Text | ✓ | ✓ | 1M | lowhighmax | usage | €0.20 | €0.60 | €0.03 | EssentialProfessionalAgentic | |
| Qwen 3.8 27B FP8 |  Alibaba | Hosted on T-Cloud Public — Telekom's sovereign infrastructure in Germany. | Text, Image | Text | ✓ | ✓ | 256K | nonelowmediumxhigh | completion_tokens_details | €0.50 | €2.50 | €0.075 | EssentialProfessionalAgentic | |
| GPT-5 Codex |  OpenAI | Hosted on Microsoft Azure (EU regions). | Text, Image | Text | ✓ | ✓ | 400K | - | - | €1.10 | €8.70 | n/a | EssentialProfessionalAgentic | |
| No models match the current filters. | ||||||||||||||