A reactive notebook for Python — run reproducible experiments, query with SQL, execute as a script, deploy as an app, and version with git. Stored as pure Python. All in a modern, AI-native editor.
Python artificial-intelligencedagdata-sciencedata-visualizationdataflowdeveloper-toolsmachine-learningnotebookspipelinepythonreactivesqlweb-app
WandB models updated (#10040)
vincent d warmerdambc14b190a7ec
diff --git a/packages/llm-info/data/models.yml b/packages/llm-info/data/models.ymlindex 36fd33acf5..02d728f419 100644--- a/packages/llm-info/data/models.yml+++ b/packages/llm-info/data/models.yml@@ -622,21 +622,51 @@ ollama:release_date: 2026-02-15wandb:- - name: GLM-5.1+ - name: DeepSeek V4-Flash+ model: deepseek-ai/DeepSeek-V4-Flash+ description: "DeepSeek V4-Flash is an MoE model with 1M context length great for coding, reasoning, and agentic workloads"+ roles: [chat, edit]+ capabilities: [thinking, tool_calling]+ input_types: [text]+ output_types: [text]+ release_date: 2026-04-24+ cost: {input: 0.14, output: 0.28}++ - name: DeepSeek V4-Pro+ model: deepseek-ai/DeepSeek-V4-Pro+ description: "DeepSeek V4-Pro is a 1.6T-parameter MoE model with 49B active parameters excelling at advanced reasoning, coding, and complex agentic workloads"+ roles: [chat, edit]+ capabilities: [thinking, tool_calling]+ input_types: [text]+ output_types: [text]+ release_date: 2026-04-24+ cost: {input: 1.74, output: 3.46}++ - name: Z.AI GLM 5.2+ model: zai-org/GLM-5.2+ description: "GLM-5.2 is a Mixture-of-Experts language model featuring 40 billion activated parameters and a total of 744 billion parameters"+ roles: [chat, edit]+ capabilities: [thinking, tool_calling]+ input_types: [text]+ output_types: [text]+ release_date: 2026-06-13+ cost: {input: 1.39, output: 4.4}++ - name: Z.AI GLM-5.1model: zai-org/GLM-5.1description: "GLM-5.1 delivers a major leap in coding capability, with particularly significant gains in handling long-horizon tasks"roles: [chat, edit]capabilities: [thinking, tool_calling]input_types: [text]output_types: [text]- release_date: 2026-03-27+ release_date: 2026-04-07cost: {input: 1.4, output: 4.4}- name: NVIDIA Nemotron 3 Super 120Bmodel: nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-FP8description: "120B-parameter open hybrid MoE model, activating just 12B parameters for maximum compute efficiency and accuracy in complex multi-agent applications"roles: [chat, edit]- capabilities: [tool_calling]+ capabilities: [thinking, tool_calling]input_types: [text]output_types: [text]release_date: 2026-03-11@@ -646,81 +676,81 @@ wandb:model: MiniMaxAI/MiniMax-M2.5description: "SOTA large language model designed for real-world productivity"roles: [chat, edit]- capabilities: [tool_calling]+ capabilities: [thinking, tool_calling]input_types: [text]output_types: [text]release_date: 2026-02-12cost: {input: 0.3, output: 1.2}- - name: GLM 5- model: zai-org/GLM-5-FP8- description: "Z.ai's flagship open-source foundation model engineered for complex systems design and long-horizon agent workflows"+ - name: Moonshot AI Kimi K2.7 Code+ model: moonshotai/Kimi-K2.7-Code+ description: "Kimi K2.7 Code is a 1T-parameter MoE model with 32B active parameters purpose-built for long-horizon agentic coding and software engineering"roles: [chat, edit]- capabilities: [tool_calling]+ capabilities: [thinking, tool_calling]input_types: [text]output_types: [text]- release_date: 2026-02-11- cost: {input: 1, output: 3.2}+ release_date: 2026-06-12+ cost: {input: 0.94, output: 4}- - name: Kimi K2.5- model: moonshotai/Kimi-K2.5- description: "Moonshot AI's native multimodal model, delivering state-of-the-art visual coding capability and a self-directed agent swarm paradigm"+ - name: Moonshot AI Kimi K2.6+ model: moonshotai/Kimi-K2.6+ description: "Kimi K2.6 is a multimodal Mixture-of-Experts language model featuring 32 billion activated parameters and a total of 1 trillion parameters"roles: [chat, edit]capabilities: [thinking, tool_calling]input_types: [text, image]output_types: [text]- release_date: 2026-01-27- cost: {input: 0.5, output: 2.85}-- - name: DeepSeek V3.1- model: deepseek-ai/DeepSeek-V3.1- description: "Large hybrid reasoning model (671B parameters, 37B active) that supports both thinking and non-thinking modes via prompt templates"- roles: [chat, edit]- capabilities: [tool_calling]- input_types: [text]- output_types: [text]- release_date: 2025-08-21- cost: {input: 0.55, output: 1.65}+ release_date: 2026-04-20+ cost: {input: 0.95, output: 4}- name: gpt-oss-120bmodel: openai/gpt-oss-120bdescription: "Open-weight, 117B-parameter Mixture-of-Experts (MoE) language model from OpenAI designed for high-reasoning, agentic, and general-purpose production use cases"roles: [chat, edit]- capabilities: [tool_calling]+ capabilities: [thinking, tool_calling]input_types: [text]output_types: [text]release_date: 2025-08-05- cost: {input: 0.15, output: 0.6}+ cost: {input: 0.04, output: 0.14}- name: gpt-oss-20bmodel: openai/gpt-oss-20bdescription: "Open-weight 21B parameter model released by OpenAI under the Apache 2.0 license"roles: [chat, edit]- capabilities: [tool_calling]+ capabilities: [thinking, tool_calling]input_types: [text]output_types: [text]release_date: 2025-08-05- cost: {input: 0.05, output: 0.2}+ cost: {input: 0.03, output: 0.13}- - name: Qwen3 30B A3B Instruct 2507- model: Qwen/Qwen3-30B-A3B-Instruct-2507- description: "30.5B-parameter mixture-of-experts language model from Qwen, with 3.3B active parameters per inference"+ - name: Qwen3.6 35B A3B+ model: Qwen/Qwen3.6-35B-A3B+ description: "Qwen3.6-35B-A3B is an MoE multimodal model with 262K context optimized for agentic coding workflows"roles: [chat, edit]- capabilities: [tool_calling]- input_types: [text]+ capabilities: [thinking, tool_calling]+ input_types: [text, image]output_types: [text]- release_date: 2025-07-29- cost: {input: 0.1, output: 0.3}+ release_date: 2026-04-16+ cost: {input: 0.25, output: 1.25}++ - name: Qwen3.6 27B+ model: Qwen/Qwen3.6-27B+ description: "Qwen3.6-27B is a 27B dense multimodal model with 262K context built for flagship-level agentic coding"+ roles: [chat, edit]+ capabilities: [thinking, tool_calling]+ input_types: [text, image]+ output_types: [text]+ release_date: 2026-04-22+ cost: {input: 0.6, output: 3.6}- - name: Qwen3-235B-A22B-Thinking-2507- model: Qwen/Qwen3-235B-A22B-Thinking-2507- description: "High-performance, open-weight Mixture-of-Experts (MoE) language model optimized for complex reasoning tasks"+ - name: JetBrains Mellum2 12B A2.5B+ model: JetBrains/Mellum2-12B-A2.5B-Instruct+ description: "Mellum2-12B-A2.5B-Instruct is a fast MoE model with 131K context built for coding, tool use, and low-latency AI workflows"roles: [chat, edit]capabilities: [thinking, tool_calling]input_types: [text]output_types: [text]- release_date: 2025-07-25- cost: {input: 0.1, output: 0.1}+ release_date: 2026-06-01+ cost: {input: 0.05, output: 0.1}opencode-go:- name: DeepSeek V4 Flash