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.
4 is the maximum. Remove one to add another.
| Attribute | Grok 3 Betagrok-3-beta | Hy4 previewhy4-preview | Muse Spark 1.3muse-spark-1.3 | Qwen3.8 27Bqwen3.8-27b |
|---|---|---|---|---|
| Pricing | ||||
| Input | $3.00 / 1M | $0.834 / 1M | $1.25 / 1M | $0.45 / 1M |
| Output | $15.00 / 1M | $2.50 / 1M | $4.25 / 1M | $3.20 / 1M |
| Cache Write (5m) | $3.00 / 1M | $0.834 / 1M | $1.25 / 1M | $0.45 / 1M |
| Cache Write (1h) | $3.00 / 1M | $0.834 / 1M | $1.25 / 1M | $0.45 / 1M |
| Cache Read | $3.00 / 1M | $0.834 / 1M | $1.25 / 1M | $0.45 / 1M |
| Web Search | $0 / 1M | $0 / 1M | $0 / 1M | $0 / 1M |
| Context | ||||
| Max context | 131.1K | 1M | 1M | 262K |
| Max output | N/A | N/A | N/A | N/A |
| Capabilities | ||||
| Vision | No | Yes | Yes | Yes |
| Function Calling | No | Yes | Yes | Yes |
| JSON Mode | No | Yes | Yes | Yes |
| Streaming | No | Yes | Yes | Yes |
| Catalogue | ||||
| Provider | xAI | Tencent | Meta | Alibaba |
| Category | chat | chat | chat | chat |
| Charge type | Pay As You Go | Pay As You Go | Pay As You Go | Pay As You Go |
| Released | — | — | — | — |
| Description | ||||
| Summary | — | Tencent Hy4 Preview is a Mixture-of-Experts (MoE) model from Tencent, featuring 770B total parameters with 49B activated per token. It is designed for coding agents, complex tool-driven workflows, and professional productivity tasks that require strong planning and reliable execution. Optimized for context continuity and sustained multi-step work, Hy4 Preview is well suited for long-horizon coding, agentic automation, tool orchestration, and complex real-world workflows. | Muse Spark 1.3 is Meta's multimodal reasoning model designed for long-running agentic, multi-agent, and coding workflows. It maintains context and information across extended tasks, enabling reliable execution in complex, multi-step environments. The model is optimized to resolve conflicting information, seek clarification or confirmation when necessary, and execute concisely, making it well suited for autonomous agents, collaborative multi-agent systems, and long-horizon software engineering workflows. | Qwen3.8 27B is an open-weight dense vision-language model from Qwen, designed for coding, professional knowledge work, research, and multimodal interaction. It combines strong text and visual understanding with capabilities optimized for sustained, real-world agentic tasks. The model supports flexible thinking modes that can be enabled for deeper reasoning or disabled for faster execution, making it well suited for long-running agents, multimodal workflows, coding assistants, and cost-conscious self-hosted deployments. |