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.
- Qwen3.8 2.4T A95BAlibabaRemove
- DeepSeek V4.1 FlashDeepSeekRemove
- Grok 4.7SpaceXAIRemove
- GPT-6 SolOpenAIRemove
4 is the maximum. Remove one to add another.
| Attribute | Qwen3.8 2.4T A95Bqwen3.8-2.4t-a95b | DeepSeek V4.1 Flashdeepseek-v4.1-flash | Grok 4.7grok-4.7 | GPT-6 Solgpt-6-sol |
|---|---|---|---|---|
| Pricing | ||||
| Input | $1.80 / 1M | $0.30 / 1M | $1.60 / 1M | $2.00 / 1M |
| Output | $5.40 / 1M | $1.20 / 1M | $4.80 / 1M | $10.00 / 1M |
| Cache Write (5m) | $1.80 / 1M | $0.30 / 1M | $1.60 / 1M | $2.00 / 1M |
| Cache Write (1h) | $1.80 / 1M | $0.30 / 1M | $1.60 / 1M | $2.00 / 1M |
| Cache Read | $1.80 / 1M | $0.30 / 1M | $1.60 / 1M | $2.00 / 1M |
| Web Search | $0 / 1M | $0 / 1M | $0 / 1M | $0 / 1M |
| Context | ||||
| Max context | 262K | 1M | 500K | 1.1M |
| Max output | N/A | N/A | N/A | N/A |
| Capabilities | ||||
| Vision | Yes | Yes | Yes | Yes |
| Function Calling | Yes | Yes | Yes | Yes |
| JSON Mode | Yes | Yes | Yes | Yes |
| Streaming | Yes | Yes | Yes | Yes |
| Catalogue | ||||
| Provider | Alibaba | DeepSeek | SpaceXAI | OpenAI |
| 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 | Qwen3.8 2.4T A95B is Qwen's open-weight sparse Mixture-of-Experts (MoE) model and the open-weight counterpart to Qwen3.8 Max. It features 2.4T total parameters with 95B activated per token, combining frontier-scale capacity with efficient sparse inference. Designed for coding, research, complex reasoning, and agentic workflows, the model is well suited for demanding long-horizon tasks and advanced autonomous systems while providing the flexibility and customization benefits of open weights. | DeepSeek V4.1 Flash is a cost-efficient sparse Mixture-of-Experts (MoE) model in DeepSeek's V4.1 family, optimized for coding, reasoning, and agentic workflows. Despite its efficiency-focused positioning, DeepSeek reports that it surpasses the previous V4 Pro in performance, inference speed, and overall task completion time. The model is particularly strong at long-horizon, multi-step execution, making it well suited for coding agents, complex problem solving, and autonomous workflows that must reliably carry tasks through to completion. | Grok 4.7 is SpaceXAI's flagship model for coding, agentic workflows, and professional knowledge work, succeeding Grok 4.6. It is particularly strong at long-running software engineering, self-verification, and long-context execution, while improving capabilities in document drafting, presentations, and other professional tasks. Trained with extended reinforcement learning focused on multi-hour problems, Grok 4.7 is optimized for sustained, complex task execution and natively supports the Grok Bot harness for conversational workflows. It also introduces an enhanced safeguard stack designed to combine strong jailbreak resistance with low refusal rates for legitimate technical work. Reported benchmark results use xhigh reasoning effort. | GPT-6 Sol is OpenAI's cost-efficient high-end model in the GPT-6 series, positioned between the flagship GPT-6 Astra and the fast GPT-6 Luna tier. It is designed for professional knowledge work, agentic coding, business workflow automation, and computer-use tasks, with particular strength in long-horizon software engineering across real-world codebases. GPT-6 Sol approaches Astra-level factual reliability at a significantly lower cost, while sharing its clear and concise communication style. This balance of capability, reliability, and efficiency makes it well suited for production agents, complex engineering workflows, and scalable professional applications |