Neither wins outright: Anthropic: Claude Sonnet 4.6 (batch) is the pick when you need image input, Qwen: Qwen3.8 2.4T A95B (batch) when you need log-probabilities.
| Spec | Anthropic: Claude Sonnet 4.6 (batch) | Qwen: Qwen3.8 2.4T A95B (batch) |
|---|---|---|
| Provider | Anthropic | Qwen |
| Input price (per 1M tokens) | $1.50 | $2.00 |
| Output price (per 1M tokens) | $7.50 | $6.00 |
| Blended price (3:1 input/output) | $3.00 | $3.00 |
| Cheaper than … of all models we track | 21% | 21% |
| Context window | 1M tokens (~1,500 pages) | 1.0M tokens (~1,515 pages) |
| Larger context than … of all models | 69% | 79% |
| Max output per response | 128K tokens (~96,000 words) | 909K tokens (~681,800 words) |
| Input types | text, image, file | text |
| Output types | text | text |
| Free variant | No | No |
| Listed since | 2026-02-17 | 2026-08-12 |
Monthly cost at list price for five common workloads. Token counts per unit are stated so you can scale them to your own traffic. A workload that does not fit a model's context window is marked rather than priced.
| Workload (per month) | Anthropic: Claude Sonnet 4.6 (batch) | Qwen: Qwen3.8 2.4T A95B (batch) | Difference |
|---|---|---|---|
| Customer-support chatbot 50,000 conversations — each a 6-turn conversation (1,500 in / 400 out tokens) | $263 | $270 | Anthropic: Claude Sonnet 4.6 (batch) saves $7.50 |
| RAG search over documents 20,000 questions — each a question answered from ~8 retrieved passages (6,000 in / 500 out tokens) | $255 | $300 | Anthropic: Claude Sonnet 4.6 (batch) saves $45.00 |
| Coding / tool-using agent 2,000 tasks — each a multi-step task with tool calls and file context (60,000 in / 6,000 out tokens) | $270 | $312 | Anthropic: Claude Sonnet 4.6 (batch) saves $42.00 |
| Batch summarisation 10,000 documents — each a 6,000-word report summarised to a page (8,000 in / 600 out tokens) | $165 | $196 | Anthropic: Claude Sonnet 4.6 (batch) saves $31.00 |
| Long-document analysis 500 documents — each a 150-page contract or codebase read in one pass (100,000 in / 2,000 out tokens) | $82.50 | $106 | Anthropic: Claude Sonnet 4.6 (batch) saves $23.50 |
| Capability | Anthropic: Claude Sonnet 4.6 (batch) | Qwen: Qwen3.8 2.4T A95B (batch) | What it lets you do |
|---|---|---|---|
| Tool / function calling | ✓ Yes | ✓ Yes | call your APIs and run agent loops |
| Structured output (JSON schema) | ✓ Yes | ✓ Yes | return JSON that validates against your schema |
| Reasoning / thinking mode | ✓ Yes | ✓ Yes | spend extra tokens thinking before answering hard problems |
| Image input | ✓ Yes | — | read screenshots, charts and scanned pages |
| Audio input | — | — | take speech or audio directly |
| File / PDF input | ✓ Yes | — | accept documents without your own parsing step |
| Image output | — | — | generate images, not just text |
| Built-in web search | ✓ Yes | — | answer from live web results |
| Seed (reproducible sampling) | — | — | repeat a generation for tests and evals |
| Log-probabilities | — | ✓ Yes | read token confidence for classification and scoring |
From each model's published parameters and input/output types. “—” means not advertised, not proven absent.
| Date | Model | Input / 1M | Output / 1M |
|---|---|---|---|
| 2026-09-02 | Qwen: Qwen3.8 2.4T A95B (batch) | $2.00 | $6.00 |
Neither wins outright: Anthropic: Claude Sonnet 4.6 (batch) is the pick when you need image input, Qwen: Qwen3.8 2.4T A95B (batch) when you need log-probabilities. In short: Anthropic: Claude Sonnet 4.6 (batch) when you need image input, you need file / PDF input, you need built-in web search; Qwen: Qwen3.8 2.4T A95B (batch) when you need log-probabilities, you generate long outputs in one go, you want the newer model.
They cost the same: $3.00 per 1M tokens at a 3:1 input/output mix.
Anthropic: Claude Sonnet 4.6 (batch) costs $1.50 per 1M input tokens and $7.50 per 1M output tokens; Qwen: Qwen3.8 2.4T A95B (batch) costs $2.00 input and $6.00 output. Output tokens are billed separately, so long answers shift the gap.
Qwen: Qwen3.8 2.4T A95B (batch): 1.0M tokens, about 1,515 pages, against 1M (about 1,500 pages) for Anthropic: Claude Sonnet 4.6 (batch).
Tool calling — Anthropic: Claude Sonnet 4.6 (batch): yes, Qwen: Qwen3.8 2.4T A95B (batch): yes. JSON-schema structured output — Anthropic: Claude Sonnet 4.6 (batch): yes, Qwen: Qwen3.8 2.4T A95B (batch): yes.
Only Anthropic: Claude Sonnet 4.6 (batch) accepts image input; Qwen: Qwen3.8 2.4T A95B (batch) is text-only for input.
No free variant is listed for either; both are pay-per-token.
Anthropic: Claude Sonnet 4.6 (batch) can write up to 128K tokens (about 96,000 words) per response; Qwen: Qwen3.8 2.4T A95B (batch) up to 909K (about 681,800 words).
Anthropic: Claude Sonnet 4.6 (batch) was listed on 2026-02-17; Qwen: Qwen3.8 2.4T A95B (batch) on 2026-08-12.
Anthropic: Claude Sonnet 4.6 (batch): price unchanged since we started tracking it on 2026-08-07. Qwen: Qwen3.8 2.4T A95B (batch): input price cut 20% ($2.50 → $2.00/1M), output price cut 4% ($6.25 → $6.00/1M) on 2026-09-02.
Sonnet 4.6 is Anthropic's most capable Sonnet-class model yet, with frontier performance across coding, agents, and professional work. It excels at iterative development, complex codebase navigation, end-to-end project management with...
Full Anthropic: Claude Sonnet 4.6 (batch) specs →Qwen3.8 2.4T A95B is an open-weight sparse mixture-of-experts model from Qwen and the open-weight variant of [Qwen3.8 Max](/qwen/qwen3.8-max), with 95 billion active parameters out of 2.4 trillion total. It is...
Full Qwen: Qwen3.8 2.4T A95B (batch) specs →