Skip to content

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.

WandB models updated (#10040)

vincent d warmerdambc14b190a7ec
diff --git a/packages/llm-info/data/models.yml b/packages/llm-info/data/models.yml
index 36fd33acf5..02d728f419 100644
--- a/packages/llm-info/data/models.yml
+++ b/packages/llm-info/data/models.yml
@@ -622,21 +622,51 @@ ollama:
release_date: 2026-02-15
wandb:
- - 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-5.1
model: zai-org/GLM-5.1
description: "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: 2026-04-07
cost: {input: 1.4, output: 4.4}
- name: NVIDIA Nemotron 3 Super 120B
model: nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-FP8
description: "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-M2.5
description: "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: 2026-02-12
cost: {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-120b
model: openai/gpt-oss-120b
description: "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-20b
model: openai/gpt-oss-20b
description: "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