feat(wafer/glm5.2-fast): add new models [bot]

[models-bot]

Auto-generated by model-addition-agent for wafer/glm5.2-fast.

---

Note

Low Risk
Additive catalog-only YAML with no runtime or auth changes; misconfigured costs or limits would only affect routing/billing metadata for this model.

Overview
Adds a new Wafer provider definition for glm5.2-fast, registering the model as active serverless chat with thinking enabled.

The entry specifies token pricing (including cache-read input), a 1M context window, text-only I/O, and capabilities such as function calling, structured output, tool choice, and prompt caching. supportedModes is chat-only (unlike GLM-5.2, which also lists responses).

^{Reviewed by Cursor Bugbot for commit 84dd05c. Bugbot is set up for automated code reviews on this repo. Configure here.}

[github-actions]

/test-models

[harshiv-26]

Gateway test results

• Total: 8
• Passed: 8
• Failed: 0
• Validation failed: 0
• Errored: 0
• Skipped: 0
• Success rate: 100.0%

Provider	Model	Scenarios
wafer	glm5.2-fast	success: tool-call, params, structured-output, params:stream, tool-call:stream, structured-output:stream, reasoning:stream, reasoning

Successes (8)

wafer/glm5.2-fast — tool-call (success)

Output

Function: get_weather
Arguments: {"location": "London"}
VALIDATION: tool-call SUCCESS

wafer/glm5.2-fast — params (success)

Output

The capital of France is Paris.

wafer/glm5.2-fast — structured-output (success)

Output

{"name":"Science Fair","date":"Friday","participants":["Alice","Bob"]}
VALIDATION: structured-output SUCCESS

wafer/glm5.2-fast — params:stream (success)

Output

The capital of France is Paris.

wafer/glm5.2-fast — tool-call:stream (success)

Output

{"location": "London"}
VALIDATION: tool-call stream SUCCESS

wafer/glm5.2-fast — structured-output:stream (success)

Output

{"name": "Science Fair", "date": "Friday", "participants": ["Alice", "Bob"]}
VALIDATION: structured-output stream SUCCESS

wafer/glm5.2-fast — reasoning:stream (success)

Output

To calculate the expression **3^3^3^3**, we first need to understand the order of operations. In mathematics, exponentiation is **right-associative** 
... (truncated, 2006 chars omitted)

wafer/glm5.2-fast — reasoning (success)

Output

To calculate $3^{3^{3^{3}}}$, we first need to understand the order of operations for exponentiation. In mathematics, stacked exponents are evaluated 
... (truncated, 1979 chars omitted)

[cursor]

Cursor Bugbot has reviewed your changes using default effort and found 2 potential issues.

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit 84dd05c. Configure here.}

[cursor]

Incomplete model feature flags

Medium Severity

The new glm5.2-fast entry omits system_messages and json_output even though the same-weights twin z-ai/glm-5.2-fast and sibling GLM-5.2 both declare them. Registry consumers filtering on those capabilities will incorrectly exclude this model.

 

^{Reviewed by Cursor Bugbot for commit 84dd05c. Configure here.}

[cursor]

Likely wrong model identifier

High Severity

The model value glm5.2-fast drops the hyphen used everywhere else for this model, including this file's own Vercel source URL, the existing twin z-ai/glm-5.2-fast, and Wafer's sibling IDs like GLM-5.2. A mismatched provider API id would make requests using this registry entry fail.

 

^{Reviewed by Cursor Bugbot for commit 84dd05c. Configure here.}