Compare models
Put up to 4 models beside each other — token prices, context windows, capabilities and provider, from the same catalogue the model pages read.
| Attribute | Qwen3 VL 235B A22B Thinkingqwen3-vl-235b-a22b-thinking | Muse Spark Openmuse-spark-open | Gemini 4 Carbongemini-4-carbon |
|---|---|---|---|
| Pricing | |||
| Input | $0.30 / 1M | $1.25 / 1M | $2.00 / 1M |
| Output | $3.00 / 1M | $4.25 / 1M | $10.00 / 1M |
| Cache Write (5m) | $0.30 / 1M | $1.25 / 1M | $2.00 / 1M |
| Cache Write (1h) | $0.30 / 1M | $1.25 / 1M | $2.00 / 1M |
| Cache Read | $0.30 / 1M | $1.25 / 1M | $2.00 / 1M |
| Web Search | $0 / 1M | $0 / 1M | $0 / 1M |
| Context | |||
| Max context | 131.1K | 1M | 1M |
| Max output | N/A | N/A | N/A |
| Capabilities | |||
| Vision | No | Yes | Yes |
| Function Calling | Yes | Yes | Yes |
| JSON Mode | Yes | Yes | Yes |
| Streaming | Yes | Yes | Yes |
| Catalogue | |||
| Provider | Alibaba | Meta | |
| Category | chat | chat | chat |
| Charge type | Pay As You Go | Pay As You Go | Pay As You Go |
| Released | — | — | — |
| Description | |||
| Summary | Qwen3-VL-235B-A22B Thinking is a powerful multimodal model that combines advanced text generation with strong image and video understanding, optimized for STEM and math reasoning. It offers robust perception, spatial grounding, and long-form visual comprehension, and supports agent-style interactions such as multi-image dialogue, video timeline alignment, GUI control, and visual-to-code workflows. With competitive benchmark results and strong text-only ability, it’s suited for production uses like document AI, OCR, UI assistance, spatial tasks, and vision-language research. | Muse Spark Open is an open-weight, coding-optimized model designed for agentic software engineering workflows. It delivers strong first-attempt accuracy, reliable tool calling, and deep codebase understanding, enabling effective development across real-world repositories. Optimized for autonomous coding agents, complex code navigation, and tool-driven development, Muse Spark Open is well suited for software engineering automation, debugging, code generation, and multi-step coding workflows where accuracy and reliable execution are essential. | Gemini 4 Carbon is Google's frontier multimodal reasoning model, designed for advanced coding, agentic workflows, and complex multi-step problem solving. It supports text, image, audio, video, and file inputs, enabling integrated reasoning and analysis across diverse modalities. Optimized for long-horizon execution and demanding professional workloads, Gemini 4 Carbon supports up to 1M output tokens through continuation and configurable reasoning effort levels (low, medium, high) to balance reasoning depth, latency, and cost. It is well suited for autonomous agents, software engineering, multimodal analysis, and large-scale workflow automation. |