Updates model TOMLs for the deepinfra sync target.

Notices

Deep Infra

• 140 Deep Infra models were not created because they lacked provider-agnostic metadata to inherit (no models/ entry) and the API does not supply the required curated fields, or because they are already deprecated.
• Skipped remote IDs: moonshotai/Kimi-K2-Thinking, Qwen/Qwen3.5-2B, Qwen/QwQ-32B, nvidia/Nemotron-Content-Safety-3.5, allenai/olmOCR-2-7B-1025, Austism/chronos-hermes-13b-v2, meta-llama/Meta-Llama-3-70B-Instruct, Qwen/Qwen3-VL-30B-A3B-Instruct, nvidia/Llama-3.1-Nemotron-70B-Instruct, mattshumer/Reflection-Llama-3.1-70B, mistralai/Mixtral-8x22B-Instruct-v0.1, zai-org/GLM-4.5, Qwen/Qwen3-VL-235B-A22B-Thinking, moonshotai/Kimi-K2-Instruct, microsoft/phi-4, meta-llama/Llama-Guard-4-12B, Qwen/Qwen3-VL-30B-A3B-Thinking, meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo, deepseek-ai/DeepSeek-V3-0324, stepfun-ai/Step-3.5-Flash, deepseek-ai/DeepSeek-V3.2-Exp, mistralai/Mistral-Small-3.1-24B-Instruct-2503, moonshotai/Kimi-K2-Instruct-0905, Qwen/Qwen3-Coder-30B-A3B-Instruct, Qwen/Qwen3-Coder-480B-A35B-Instruct, mistralai/Mistral-7B-Instruct-v0.3, meta-llama/Llama-Guard-3-8B, meta-llama/Llama-4-Maverick-17B-128E-Instruct-Turbo, mistralai/Mixtral-8x7B-Instruct-v0.1, deepinfra/airoboros-70b, Sao10K/L3.1-70B-Euryale-v2.2, google/gemma-4-31B-it-turbo, PaddlePaddle/PaddleOCR-VL-0.9B, deepseek-ai/DeepSeek-V3-0324-Turbo, MiniMaxAI/MiniMax-M2, meta-llama/Llama-3.2-3B-Instruct, google/gemma-2-27b-it, allenai/olmOCR-7B-0825, Qwen/Qwen3-VL-4B-Instruct, allenai/Olmo-3.1-32B-Instruct, deepseek-ai/DeepSeek-R1-0528-Turbo, mistralai/Mistral-7B-Instruct-v0.2, Qwen/Qwen3-235B-A22B-Instruct-2507, KoboldAI/LLaMA2-13B-Tiefighter, google/gemma-1.1-7b-it, meta-llama/Meta-Llama-3.1-405B-Instruct, Qwen/Qwen2.5-VL-32B-Instruct, microsoft/Phi-4-multimodal-instruct, meta-llama/Llama-3.2-1B-Instruct, ByteDance/Seed-2.0-code, Gryphe/MythoMax-L2-13b-turbo, MiniMaxAI/MiniMax-M2.7-Turbo, google/gemma-2-9b-it, Phind/Phind-CodeLlama-34B-v2, openchat/openchat-3.6-8b, mistralai/Devstral-Small-2507, ByteDance/Seed-2.0-mini, lizpreciatior/lzlv_70b_fp16_hf, microsoft/Phi-3-medium-4k-instruct, moonshotai/Kimi-K2.5-Turbo, nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16, nvidia/NVIDIA-Nemotron-Nano-12B-v2-VL, mistralai/Mistral-Small-24B-Instruct-2501, allenai/olmOCR-2, MiniMaxAI/MiniMax-M2.1, Sao10K/L3-70B-Euryale-v2.1, Sao10K/L3-8B-Lunaris-v1, stepfun-ai/Step-3.7-Flash, Qwen/Qwen2.5-7B-Instruct, google/codegemma-7b-it, meta-llama/Llama-2-70b-chat-hf, zai-org/GLM-4.5-Air, openbmb/MiniCPM-Llama3-V-2_5, Qwen/Qwen3.5-4B, deepseek-ai/DeepSeek-V3.1, meta-llama/Meta-Llama-3-8B-Instruct, Qwen/Qwen3.5-0.8B, mistralai/Mistral-7B-Instruct-v0.1, ByteDance/Seed-2.0-pro, nvidia/NVIDIA-Nemotron-3-Super-120B-A12B, bigcode/starcoder2-15b-instruct-v0.1, Qwen/Qwen3-VL-8B-Instruct, Sao10K/L3-8B-Lunaris-v1-Turbo, mistralai/Mistral-Nemo-Instruct-2407, meta-llama/Meta-Llama-3.1-70B-Instruct, Qwen/Qwen2.5-72B-Instruct, Sao10K/L3.3-70B-Euryale-v2.3, mistralai/Mistral-Small-3.2-24B-Instruct-2506, Qwen/Qwen2-7B-Instruct, Qwen/Qwen2.5-Coder-32B-Instruct, Qwen/Qwen3-14B, microsoft/WizardLM-2-8x22B, zai-org/GLM-4.6V, Qwen/QVQ-72B-Preview, nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B, deepreinforce-ai/Ornith-1.0-35B, meta-llama/Meta-Llama-3.1-8B-Instruct, mistralai/Devstral-Small-2505, microsoft/phi-4-reasoning-plus, NovaSky-AI/Sky-T1-32B-Preview, deepseek-ai/DeepSeek-R1-Distill-Qwen-32B, tencent/Hy3, Qwen/Qwen3-235B-A22B-Thinking-2507, cognitivecomputations/dolphin-2.9.1-llama-3-70b, meta-llama/Llama-3.3-70B-Instruct, deepseek-ai/DeepSeek-R1, deepseek-ai/DeepSeek-V3, Gryphe/MythoMax-L2-13b, Qwen/Qwen3-Max-Thinking, cognitivecomputations/dolphin-2.6-mixtral-8x7b, Qwen/Qwen3-VL-235B-A22B-Instruct, nvidia/NVIDIA-Nemotron-Nano-9B-v2, Qwen/Qwen3-235B-A22B, microsoft/WizardLM-2-7B, meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo, deepseek-ai/DeepSeek-R1-Turbo, Qwen/Qwen3-VL-4B-Thinking, allenai/olmOCR-7B-0725-FP8, Qwen/Qwen3-30B-A3B, deepseek-ai/DeepSeek-V3.1-Terminus, meta-llama/Llama-2-13b-chat-hf, NousResearch/Hermes-3-Llama-3.1-70B, deepseek-ai/DeepSeek-Prover-V2-671B, google/gemma-3-12b-it, NousResearch/Hermes-3-Llama-3.1-405B, allenai/olmOCR-7B-1025, deepseek-ai/DeepSeek-OCR, openchat/openchat_3.5, Qwen/QwQ-32B-Preview, meta-llama/Llama-3.2-90B-Vision-Instruct, Qwen/Qwen2-72B-Instruct, Qwen/Qwen3-VL-8B-Thinking, meta-llama/Llama-3.2-11B-Vision-Instruct, google/gemma-3-4b-it, google/gemma-3-27b-it, Qwen/Qwen2.5-Coder-7B, ByteDance/Seed-1.8, deepseek-ai/DeepSeek-R1-Distill-Llama-70B, nvidia/Nemotron-4-340B-Instruct, openai/gpt-oss-120b-Turbo
• Add a models/<provider>/<model>.toml entry (or a full provider TOML) to include them in the next sync.

This PR was created automatically by the daily model sync workflow.