Feat: Add Nano-GPT provider models

[b3nw]

Adds Nano-GPT as a new provider with model definitions.

Provider Details

• Name: NanoGPT
• API: https://nano-gpt.com/api/v1
• Docs: https://docs.nano-gpt.com

Models

Includes models from DeepSeek, Qwen, Mistral, Meta/Llama, Moonshot (Kimi), MiniMax, NousResearch, NVIDIA, OpenAI OSS, and Zhipu (GLM).

[Copilot]

Pull request overview

This PR adds NanoGPT as a new AI model provider with 20 model definitions spanning multiple organizations including DeepSeek, Qwen, Mistral, Meta/Llama, Moonshot (Kimi), MiniMax, NousResearch, NVIDIA, OpenAI OSS, and Zhipu (GLM). The provider configuration uses an OpenAI-compatible SDK and specifies the API endpoint and authentication requirements.

Key Changes

• Added NanoGPT provider configuration with API endpoint https://nano-gpt.com/api/v1
• Defined 20 models across 10 different AI model families with capabilities including reasoning, tool calling, and structured output
• Models include both standard and thinking/reasoning variants with context windows ranging from 32K to 262K tokens

Reviewed changes

Copilot reviewed 22 out of 22 changed files in this pull request and generated 7 comments.

Show a summary per file

File	Description
providers/nano-gpt/provider.toml	Provider configuration defining NanoGPT with environment variable, npm package, API endpoint, and documentation URL
providers/nano-gpt/models/deepseek/deepseek-r1.toml	DeepSeek R1 reasoning model with 128K context window
providers/nano-gpt/models/deepseek/deepseek-v3.2:thinking.toml	DeepSeek V3.2 thinking variant model
providers/nano-gpt/models/meta-llama/llama-3.3-70b-instruct.toml	Meta Llama 3.3 70B instruction-tuned model
providers/nano-gpt/models/meta-llama/llama-4-maverick.toml	Meta Llama 4 Maverick model with 128K context
providers/nano-gpt/models/minimax/minimax-m2.1.toml	MiniMax M2.1 model definition
providers/nano-gpt/models/mistralai/devstral-2-123b-instruct-2512.toml	Mistral Devstral 2 123B instruction model
providers/nano-gpt/models/mistralai/ministral-14b-instruct-2512.toml	Mistral Ministral 14B instruction model
providers/nano-gpt/models/mistralai/mistral-large-3-675b-instruct-2512.toml	Mistral Large 3 675B instruction model
providers/nano-gpt/models/moonshotai/kimi-k2-instruct.toml	Moonshot Kimi K2 instruction model with 131K context
providers/nano-gpt/models/moonshotai/kimi-k2-thinking.toml	Moonshot Kimi K2 thinking/reasoning variant
providers/nano-gpt/models/nousresearch/hermes-4-405b:thinking.toml	NousResearch Hermes 4 405B thinking model
providers/nano-gpt/models/nvidia/llama-3_3-nemotron-super-49b-v1_5.toml	NVIDIA Llama 3.3 Nemotron Super 49B model
providers/nano-gpt/models/openai/gpt-oss-120b.toml	OpenAI OSS GPT 120B open source model
providers/nano-gpt/models/qwen/qwen3-235b-a22b-thinking-2507.toml	Qwen3 235B thinking model with 262K context
providers/nano-gpt/models/qwen/qwen3-coder.toml	Qwen3 Coder model specialized for coding tasks
providers/nano-gpt/models/z-ai/glm-4.6.toml	Zhipu GLM 4.6 model with 200K context
providers/nano-gpt/models/z-ai/glm-4.6:thinking.toml	Zhipu GLM 4.6 thinking variant
providers/nano-gpt/models/zai-org/glm-4.5-air.toml	Zhipu GLM 4.5 Air lightweight model
providers/nano-gpt/models/zai-org/glm-4.5-air:thinking.toml	Zhipu GLM 4.5 Air thinking variant
providers/nano-gpt/models/zai-org/glm-4.7.toml	Zhipu GLM 4.7 model with reasoning support
providers/nano-gpt/models/zai-org/glm-4.7:thinking.toml	Zhipu GLM 4.7 thinking variant

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Pull request overview

Copilot reviewed 22 out of 22 changed files in this pull request and generated 5 comments.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[b3nw]

updated