Deepseek V4.1 Flash 1017 Lab: DeepSeek-AI DeepSeek V4.1 Flash 1017 is DeepSeek's open-weight multimodal MoE model with 552B backbone parameters and a 1M-token context window. It accepts text and images and improves reasoning, agentic coding, and visual analysis while completing tasks faster with shorter outputs.
Features
ToolsReasoningVisionJSON
Limits
- Context
- 1M
- Max output
- 82K
|
TL5
0.25 €
0.28 $
TL1
0.20 €
0.22 $
TL0
0.28 €
0.31 $
|
TL5
0.10 €
0.11 $
TL1
0.018 €
0.020 $
TL0
0.03 €
0.03 $
|
TL5
1.00 €
1.10 $
TL1
0.70 €
0.77 $
TL0
1.10 €
1.21 $
|
TL5
1M
TL1
—
TL0
—
|
82K
|
DESGUS
|
ToolsReasoningVisionJSON
|
TL5
TL1
TL0
|
Kimi K3 Lab: Moonshot AI Kimi K3 is Moonshot AI's open-weight multimodal Mixture-of-Experts model with 2.8T total parameters, 104B active per token, and a 1M-token context window. It accepts text and images for long-horizon coding, visual analysis, deep research, and tool-driven agents.
Features
ToolsReasoningVisionJSON
Limits
- Context
- —
- Max output
- —
|
TL1
3.00 €
3.30 $
TL0
2.00 €
2.20 $
|
TL1
0.30 €
0.33 $
TL0
0.25 €
0.28 $
|
TL1
15.00 €
16.52 $
TL0
12.00 €
13.22 $
|
—
|
—
|
SGUS
|
ToolsReasoningVisionJSON
|
TL1
TL0
|
Qwen 3.6 27B Lab: Alibaba Cloud Qwen3.6-27B is Alibaba's open-weight, 27B-parameter dense vision-language model with 262K native context, extendable to about 1M tokens. It accepts text, images, and video for agentic coding, visual document analysis, and tool-based workflows with optional thinking.
Features
ToolsReasoningVisionJSON
Limits
- Context
- —
- Max output
- —
|
TL3
0.25 €
0.28 $
TL0
0.24 €
0.26 $
|
TL3
0.075 €
0.083 $
TL0
0.05 €
0.06 $
|
TL3
1.50 €
1.65 $
TL0
1.85 €
2.04 $
|
—
|
—
|
FRUS
|
ToolsReasoningVisionJSON
|
TL3
TL0
|
Qwen 3.8 Max 0902 Lab: Alibaba Cloud Qwen3.8-Flash is Alibaba's production multimodal MoE model with a 125B-parameter backbone, 6B active per token, and 51B additional n-gram embedding parameters. It supports 1M-token context, text, image, and video inputs for coding, visual analysis, and tool-driven office agents.
Features
ToolsReasoningVisionJSON
Limits
- Context
- —
- Max output
- —
|
TL1
2.00 €
2.20 $
TL0
1.80 €
1.98 $
|
TL1
0.40 €
0.44 $
TL0
0.25 €
0.28 $
|
TL1
5.50 €
6.06 $
TL0
5.50 €
6.06 $
|
—
|
—
|
SGUS
|
ToolsReasoningVisionJSON
|
TL1
TL0
|
Deepseek V4 Pro 0813 Lab: DeepSeek-AI DeepSeek V4 Pro 0813 is DeepSeek's open-weight, text-only Mixture-of-Experts model with 1.65T parameters, 49B active per token, and a 1M-token context window. Built for complex software engineering and agentic tool use, it supports low, high, and max reasoning effort.
Features
ToolsReasoningJSON
Limits
- Context
- —
- Max output
- —
|
1.00 €
1.10 $
|
0.10 €
0.11 $
|
2.00 €
2.20 $
|
—
|
—
|
SG
|
ToolsReasoningJSON
|
TL1
|
Deepseek V4.1 Flash 1017 TL0 Lab: DeepSeek-AI DeepSeek V4.1 Flash 1017 is DeepSeek's open-weight multimodal MoE model with 552B backbone parameters and a 1M-token context window. It accepts text and images and improves reasoning, agentic coding, and visual analysis while completing tasks faster with shorter outputs.
Features
ToolsReasoningVisionJSON
Limits
- Context
- —
- Max output
- —
|
0.28 €
0.31 $
|
0.03 €
0.03 $
|
1.10 €
1.21 $
|
—
|
—
|
US
|
ToolsReasoningVisionJSON
|
TL0
|
Deepseek V4.1 Flash 1017 TL1 Lab: DeepSeek-AI DeepSeek V4.1 Flash 1017 is DeepSeek's open-weight multimodal MoE model with 552B backbone parameters and a 1M-token context window. It accepts text and images and improves reasoning, agentic coding, and visual analysis while completing tasks faster with shorter outputs.
Features
ToolsReasoningVisionJSON
Limits
- Context
- —
- Max output
- —
|
0.20 €
0.22 $
|
0.018 €
0.020 $
|
0.70 €
0.77 $
|
—
|
—
|
SG
|
ToolsReasoningVisionJSON
|
TL1
|
Deepseek V4.1 Flash 1017 TL5 Lab: DeepSeek-AI DeepSeek V4.1 Flash 1017 is DeepSeek's open-weight multimodal MoE model with 552B backbone parameters and a 1M-token context window. It accepts text and images and improves reasoning, agentic coding, and visual analysis while completing tasks faster with shorter outputs.
Features
ToolsReasoningVisionJSON
Limits
- Context
- 1M
- Max output
- 82K
Model
- Region
- DE
- Quantization
- mxfp4
|
0.25 €
0.28 $
|
0.10 €
0.11 $
|
1.00 €
1.10 $
|
1M
|
82K
|
DE
|
ToolsReasoningVisionJSON
|
TL5
|
Gemma 4 26B Lab: Google DeepMind Gemma 4 26B A4B is Google DeepMind's open-weight multimodal Mixture-of-Experts model with 26B parameters, 4B active per token, and a 256K-token context window. It handles text, images, and video frames for agentic coding, tool use, document analysis, visual reasoning, and multilingual tasks.
Features
ToolsReasoningVisionJSON
Limits
- Context
- 256K
- Max output
- 82K
Model
- Region
- DE
- Quantization
- fp8
|
0.10 €
0.11 $
|
-
-
|
0.50 €
0.55 $
|
256K
|
82K
|
DE
|
ToolsReasoningVisionJSON
|
TL5
|
Gemma 4 31B Lab: Google DeepMind Gemma 4 31B is Google DeepMind's open-weight, 30.7B-parameter dense multimodal model with a 256K-token context window. It accepts text and images, supports video analysis via frames, and generates text for complex reasoning, coding, document understanding, and tool-driven agents.
Features
ToolsReasoningVisionJSON
Limits
- Context
- —
- Max output
- —
Model
- Region
- FR
- Quantization
- fp8
|
0.15 €
0.17 $
|
0.075 €
0.083 $
|
0.40 €
0.44 $
|
—
|
—
|
FR
|
ToolsReasoningVisionJSON
|
TL3
|
GLM 5.3 Lab: Z.ai GLM-5.3 is Z.ai's open-weight, text-only Mixture-of-Experts reasoning model with 753B total and 40B active parameters, plus a 1M-token context window. Built for repository-scale coding, multi-step agents, and tool use, it offers low, high, and max reasoning effort.
Features
ToolsReasoningJSON
Limits
- Context
- 524K
- Max output
- 82K
Model
- Region
- DE
- Quantization
- mxfp4
|
1.00 €
1.10 $
|
0.25 €
0.28 $
|
3.50 €
3.85 $
|
524K
|
82K
|
DE
|
ToolsReasoningJSON
|
TL5
|
GLM 5.3 Flash Lab: Z.ai GLM-5.3-Flash is Z.ai's open-weight multimodal Mixture-of-Experts model with 320B total parameters, 18B active per token, and a 1M-token context window. It processes text and images for agentic coding, tool use, visual analysis, and long-document tasks, with adjustable reasoning effort.
Features
ToolsReasoningVisionJSON
Limits
- Context
- —
- Max output
- —
|
0.125 €
0.138 $
|
0.025 €
0.028 $
|
0.40 €
0.44 $
|
—
|
—
|
FR
|
ToolsReasoningVisionJSON
|
TL3
|
GPT-OSS 120B Lab: OpenAI GPT-OSS 120B is OpenAI's open-weight, text-only Mixture-of-Experts reasoning model with 117B total parameters, 5.1B active per token, and a 128K-token context window. It supports coding, multi-step agents, function calling, structured outputs, and low, medium, or high reasoning effort.
Features
ToolsReasoningJSON
Limits
- Context
- 128K
- Max output
- 33K
Model
- Region
- DE
- Quantization
- mxfp4
|
0.15 €
0.17 $
|
-
-
|
0.55 €
0.61 $
|
128K
|
33K
|
DE
|
ToolsReasoningJSON
|
TL5
|
Kimi K3 TL0 Lab: Moonshot AI Kimi K3 is Moonshot AI's open-weight multimodal Mixture-of-Experts model with 2.8T total parameters, 104B active per token, and a 1M-token context window. It accepts text and images for long-horizon coding, visual analysis, deep research, and tool-driven agents.
Features
ToolsReasoningVisionJSON
Limits
- Context
- —
- Max output
- —
|
2.00 €
2.20 $
|
0.25 €
0.28 $
|
12.00 €
13.22 $
|
—
|
—
|
US
|
ToolsReasoningVisionJSON
|
TL0
|
Kimi K3 TL1 Lab: Moonshot AI Kimi K3 is Moonshot AI's open-weight multimodal Mixture-of-Experts model with 2.8T total parameters, 104B active per token, and a 1M-token context window. It accepts text and images for long-horizon coding, visual analysis, deep research, and tool-driven agents.
Features
ToolsReasoningVisionJSON
Limits
- Context
- —
- Max output
- —
|
3.00 €
3.30 $
|
0.30 €
0.33 $
|
15.00 €
16.52 $
|
—
|
—
|
SG
|
ToolsReasoningVisionJSON
|
TL1
|
Kimi K3.1 Pro Lab: Moonshot AI
Features
ToolsReasoningVisionJSON
Limits
- Context
- —
- Max output
- —
|
2.00 €
2.20 $
|
0.25 €
0.28 $
|
10.00 €
11.01 $
|
—
|
—
|
SG
|
ToolsReasoningVisionJSON
|
TL1
|
Kolibri 1 Lab: Aleph Alpha Kolibri 1 is Aleph Alpha's German- and English-focused open-weight reasoning model, built in Europe with 78B total and 3.46B active parameters. It supports up to 1M tokens, tool calling, coding, and RAG for enterprise assistants, document analysis, and public-sector workflows.
Features
ToolsReasoningJSON
Limits
- Context
- 262K
- Max output
- 82K
Model
- Region
- DE
- Quantization
- fp8
|
0.15 €
0.17 $
|
0.075 €
0.083 $
|
0.55 €
0.61 $
|
262K
|
82K
|
DE
|
ToolsReasoningJSON
|
TL5
|
Mistral Small 4 Lab: Mistral AI Mistral Small 4 is Mistral AI's open-weight multimodal Mixture-of-Experts model with 119B total parameters, 6.5B active per token, and a 256K-token context window. It combines fast chat and optional deep reasoning for coding agents, document analysis, visual understanding, and tool use.
Limits
- Context
- 262K
- Max output
- 82K
Model
- Region
- DE
- Quantization
- fp8
|
0.20 €
0.22 $
|
-
-
|
0.60 €
0.66 $
|
262K
|
82K
|
DE
|
ReasoningVision
|
TL5
|
Qwen 3.6 27B TL0 Lab: Alibaba Cloud Qwen3.6-27B is Alibaba's open-weight, 27B-parameter dense vision-language model with 262K native context, extendable to about 1M tokens. It accepts text, images, and video for agentic coding, visual document analysis, and tool-based workflows with optional thinking.
Features
ToolsReasoningVisionJSON
Limits
- Context
- —
- Max output
- —
|
0.24 €
0.26 $
|
0.05 €
0.06 $
|
1.85 €
2.04 $
|
—
|
—
|
US
|
ToolsReasoningVisionJSON
|
TL0
|
Qwen 3.6 27B TL3 Lab: Alibaba Cloud Qwen3.6-27B is Alibaba's open-weight, 27B-parameter dense vision-language model with 262K native context, extendable to about 1M tokens. It accepts text, images, and video for agentic coding, visual document analysis, and tool-based workflows with optional thinking.
Features
ToolsReasoningVisionJSON
Limits
- Context
- —
- Max output
- —
|
0.25 €
0.28 $
|
0.075 €
0.083 $
|
1.50 €
1.65 $
|
—
|
—
|
FR
|
ToolsReasoningVisionJSON
|
TL3
|
Qwen 3.8 2.4T A95B Lab: Alibaba Cloud Qwen3.8 2.4T A95B is Alibaba's open-weight, text-only MoE reasoning model with 2.4T total parameters, 95B active per token, and 262K native context, extendable to 1M. It supports complex coding, research, tool calling, and long-horizon agents with adjustable reasoning effort.
Features
ToolsReasoningVisionJSON
Limits
- Context
- —
- Max output
- —
|
1.90 €
2.09 $
|
0.22 €
0.24 $
|
5.00 €
5.51 $
|
—
|
—
|
SG
|
ToolsReasoningVisionJSON
|
TL1
|
Qwen 3.8 27B Lab: Alibaba Cloud Qwen3.8 27B is Alibaba's open-weight, 27B-parameter dense vision-language model with 262K native context, extendable to 1M tokens. It processes text, images, and video for agentic coding, document and chart analysis, and multi-step workflows with configurable thinking.
Features
ToolsReasoningVisionJSON
Limits
- Context
- 262K
- Max output
- 66K
Model
- Region
- DE
- Quantization
- fp8
|
0.15 €
0.17 $
|
0.075 €
0.083 $
|
0.65 €
0.72 $
|
262K
|
66K
|
DE
|
ToolsReasoningVisionJSON
|
TL5
|
Qwen 3.8 Flash Lab: Alibaba Cloud Qwen3.8-Flash is Alibaba's production multimodal MoE model with a 125B-parameter backbone, 6B active per token, and 51B additional n-gram embedding parameters. It supports 1M-token context, text, image, and video inputs for coding, visual analysis, and tool-driven office agents.
Features
ToolsReasoningVisionJSON
Limits
- Context
- —
- Max output
- —
|
0.14 €
0.15 $
|
0.014 €
0.015 $
|
0.40 €
0.44 $
|
—
|
—
|
SG
|
ToolsReasoningVisionJSON
|
TL1
|
Qwen 3.8 Max 0902 TL0 Lab: Alibaba Cloud Qwen3.8-Flash is Alibaba's production multimodal MoE model with a 125B-parameter backbone, 6B active per token, and 51B additional n-gram embedding parameters. It supports 1M-token context, text, image, and video inputs for coding, visual analysis, and tool-driven office agents.
Features
ToolsReasoningVisionJSON
Limits
- Context
- —
- Max output
- —
|
1.80 €
1.98 $
|
0.25 €
0.28 $
|
5.50 €
6.06 $
|
—
|
—
|
US
|
ToolsReasoningVisionJSON
|
TL0
|
Qwen 3.8 Max 0902 TL1 Lab: Alibaba Cloud Qwen3.8-Flash is Alibaba's production multimodal MoE model with a 125B-parameter backbone, 6B active per token, and 51B additional n-gram embedding parameters. It supports 1M-token context, text, image, and video inputs for coding, visual analysis, and tool-driven office agents.
Features
ToolsReasoningVisionJSON
Limits
- Context
- —
- Max output
- —
|
2.00 €
2.20 $
|
0.40 €
0.44 $
|
5.50 €
6.06 $
|
—
|
—
|
SG
|
ToolsReasoningVisionJSON
|
TL1
|
Qwen 4 Max Lab: Alibaba Cloud Qwen4-Max is Alibaba's flagship multimodal reasoning model with a 1M-token context window. It handles text, images, and video, and is designed for complex coding, frontend development, professional knowledge work, and long-running agent tasks with tool calling and adjustable reasoning effort.
Features
ToolsReasoningVisionJSON
Limits
- Context
- —
- Max output
- —
|
1.50 €
1.65 $
|
0.075 €
0.083 $
|
4.50 €
4.96 $
|
—
|
—
|
SG
|
ToolsReasoningVisionJSON
|
TL1
|