fix: use kimi-k3-base-picker for GitHub Copilot

[alexneyler]

Summary

Change the GitHub Copilot Kimi model name from kimi-k3 to kimi-k3-base-picker in benchmarks/ai-gateway/providers-kimi.ts. Other providers are unchanged.

Validation

• git diff --check passed.
• Type checking could not run because npm is not available in this environment.

---

[open-cla]

Contributor License Agreement

All contributors are covered by a CLA.

[devin-ai-integration]

🔍 Devin Review: 1 flag

Not posted on this PR by your GitHub settings — view it in Devin Review. (Configure)

[github-actions]

AI Gateway Benchmark Results — OpenAI

#	Gateway	Score	Cold E2E	Warm TTFT	Tok/sec	Status
1	GitHub Copilot	90.2	565ms	522ms	69.9	20/20
2	OpenRouter	89.5	555ms	455ms	65.5	20/20
3	Concentrate AI	89.1	764ms	640ms	56.4	20/20
4	Pydantic AI Gateway	88.5	842ms	650ms	52.3	20/20
5	Vercel AI Gateway	88.2	795ms	778ms	51.4	20/20
6	OpenAI (direct)	87.9	904ms	590ms	57.1	20/20
7	LLM Gateway	87.2	1080ms	945ms	63.0	20/20
8	Neon AI Gateway	86.6	1049ms	823ms	48.0	20/20
9	LLM API	85.3	1313ms	809ms	46.3	20/20
10	ngrok AI Gateway	85.0	823ms	667ms	57.3	20/20
11	Cloudflare AI Gateway	84.4	1400ms	791ms	40.8	20/20

---

View full run · SVG available as a build artifact

[github-actions]

AI Gateway Benchmark Results — Anthropic

#	Gateway	Score	Cold E2E	Warm TTFT	Tok/sec	Status
1	Anthropic (direct)	91.8	335ms	316ms	71.2	20/20
2	ngrok AI Gateway	91.1	437ms	340ms	67.9	20/20
3	Pydantic AI Gateway	90.8	478ms	426ms	64.7	20/20
4	OpenRouter	90.4	493ms	465ms	57.1	20/20
5	GitHub Copilot	90.2	556ms	502ms	57.2	20/20
6	Vercel AI Gateway	90.0	509ms	496ms	59.6	20/20
7	LLM Gateway	89.8	569ms	483ms	60.7	20/20
8	Concentrate AI	89.7	655ms	552ms	56.9	20/20
9	Cloudflare AI Gateway	89.5	603ms	377ms	64.7	20/20
10	Neon AI Gateway	89.2	815ms	596ms	53.1	20/20
11	LLM API	88.9	889ms	559ms	55.7	20/20
12	blazerail	76.1	1043ms	963ms	41.4	18/20

---

View full run · SVG available as a build artifact

[github-actions]

AI Gateway Benchmark Results — Gemini

#	Gateway	Score	Cold E2E	Warm TTFT	Tok/sec	Status
1	OpenRouter	89.2	1547ms	1509ms	127.0	20/20
2	Neon AI Gateway	85.2	2006ms	1981ms	97.3	20/20
3	Concentrate AI	84.9	1997ms	2063ms	91.5	20/20
4	Vercel AI Gateway	59.4	6117ms	6508ms	30.4	20/20
5	LLM API	55.6	8190ms	5960ms	24.5	20/20
6	Gemini (direct)	50.1	8248ms	7326ms	26.0	19/20
7	LLM Gateway	49.7	8078ms	8070ms	24.2	20/20
8	Cloudflare AI Gateway	44.4	7659ms	7350ms	27.0	17/20
9	ngrok AI Gateway	40.7	7374ms	6891ms	26.9	14/20
10	blazerail	0.0	0ms	0ms	0.0	0/20

---

View full run · SVG available as a build artifact

[github-actions]

AI Gateway Benchmark Results — Kimi

#	Gateway	Score	Cold E2E	Warm TTFT	Tok/sec	Status
1	Neon AI Gateway	90.7	972ms	529ms	101.1	20/20
2	GitHub Copilot	90.3	374ms	367ms	113.3	20/20
3	ngrok AI Gateway	77.3	3468ms	2140ms	32.5	20/20
4	Kimi (direct)	77.3	3466ms	2161ms	33.0	20/20
5	LLM API	76.8	3679ms	2202ms	32.3	20/20
6	Vercel AI Gateway	76.7	3298ms	2278ms	31.9	20/20
7	LLM Gateway	76.6	3597ms	2304ms	32.3	20/20
8	Cloudflare AI Gateway	75.5	3537ms	2331ms	32.2	20/20
9	OpenRouter	74.5	3397ms	3464ms	31.6	20/20
10	Concentrate AI	70.7	3640ms	2843ms	30.9	19/20
11	blazerail	0.0	0ms	0ms	0.0	0/20
12	Novita	0.0	0ms	0ms	0.0	0/20

---

View full run · SVG available as a build artifact