Anthropic · Closed
Claude Sonnet 4.6 API Member pricing vs list price
anthropic/claude-sonnet-4-6
ClosedClaude Sonnet 4.6 is a closed model in Anthropic's Sonnet line, which sits between Haiku and Opus, with a 1M-token context window. Teams use it for coding agents, long-context work and prompts with strict instructions. On Bearer Club it lists at $3.00 input and $15.00 output per 1M tokens, and Members pay $2.85 and $14.25 with a Private key.
Claude Sonnet 4.6 prices per 1M tokens
Input / output, by membership tier and data policy. Every price is a percent of the provider's public list price.
| Tier | Private | Research ShareFinancial incentive notice |
|---|---|---|
| List price | $3.00 input, $15.00 output per 1M tokens | |
| Guest | Members onlyBecome a Member | Members onlyBecome a Member |
| Member | $2.85 input, $14.25 output per 1M tokens5% off list price | $2.70 input, $13.50 output per 1M tokens10% off list price |
| Executive | $2.55 input, $12.75 output per 1M tokens15% off list price | $2.40 input, $12.00 output per 1M tokens20% off list price |
Does Member pay for itself on Claude Sonnet 4.6?
Member costs $199 a year. At list price, a Member Private key saves $0.15 per 1M input tokens and $0.75 per 1M output tokens. It pays for itself at about $3,980 a year ($332 a month) of Claude Sonnet 4.6 spend at list price.
A coding agent that sends 10M tokens a day at a 3:1 input-to-output mix spends about $1,800 a month at list, so it reaches that point after about 67 days of spend. Past it, the discount is savings, and the same membership covers the other 68 models.
- Member ($199 a year) pays for itself once your Claude Sonnet 4.6 spend at list price passes $332 a month with a Private key (about 55M tokens at a 3:1 input-to-output mix) or $166 a month with Research Share.
- Executive ($1,999 a year) costs less than Member above $1,500 a month (Private keys), about 250M tokens a month.
At this volume Member · Research Share saves $43.42 a month against list price.
| Plan | Usage / mo | Fee / mo | Net vs list / mo | Pays off above |
|---|---|---|---|---|
| Member · Private | $570 | $16.58 | $13.42 | 55M tokens |
| Member · Research Share | $540 | $16.58 | $43.42 | 28M tokens |
| Executive · Private | $510 | $167 | −$76.58 | 185M tokens |
| Executive · Research Share | $480 | $167 | −$46.58 | 139M tokens |
Usage at each plan's price per token. The fee is the annual membership fee divided by 12 and covers every model, so spend on other models counts toward it too.
Specs
| Provider | Anthropic |
|---|---|
| Model type | Closed (proprietary API) |
| Context window | 1,000,000 tokens (1M) |
| Input and output | Text in, text out (chat completions) |
| Model ID | anthropic/claude-sonnet-4-6 |
| Endpoint | POST https://api.bearer.club/v1/chat/completions |
| Plans | Member, Executive |
| List price per 1M tokens | $3.00 input, $15.00 output |
List price rank: 63rd lowest input price and 62nd lowest output price of the 69 models. Its context window is larger than 33 of them.
What Claude Sonnet 4.6 is best at
Production coding agents
Coding agents plan, call tools, read errors and retry, so one task means many long requests. A 1M-token window lets an agent keep large parts of a codebase and its history in one request instead of chunking it. Bearer Club passes tools and tool_choice through where the model supports them.
Long-context document review
Contracts, technical specs and research papers rarely need splitting at 1M tokens. A 400-page document is roughly 250,000 tokens, so it fits in one request with room for your questions; reading it once costs about $0.75 at list. Very long requests take longer, so set your client timeout to match.
Complex instruction following
Prompts with nested conditions, role constraints and strict output formats are a common reason to pick a Sonnet-class model over a cheaper one. If format drift breaks your downstream code, test Claude Sonnet 4.6 against your own templates. JSON mode (response_format) passes through where the model supports it.
What 1,000 requests cost
Example request shapes, priced per 1,000 requests from the per-token prices above.
| Request (input / output tokens) | List price | Member Private | Executive Private |
|---|---|---|---|
| Short chat reply1,000 / 300 | $7.50 | $7.13 | $6.38 |
| Document summary20,000 / 1,000 | $75.00 | $71.25 | $63.75 |
| Long-context read200,000 / 2,000 | $630 | $599 | $536 |
Claude Sonnet 4.6 vs similar models
- vs Claude Haiku 4.5: Claude Haiku 4.5 costs $1.00 per 1M input tokens at list, 3× cheaper than Claude Sonnet 4.6, with a 200K-token window. It is a common pick for high-volume classification, routing and short replies. Move to Claude Sonnet 4.6 when your evals show the difference in instruction fidelity, or when inputs pass 200K tokens.
- vs Claude Opus 5.5: Claude Opus 5.5 costs $4.00 per 1M input tokens at list, 33% more than Claude Sonnet 4.6. Try it when your own evals show Claude Sonnet 4.6 failing on ambiguous, multi-step problems; otherwise start with the lower-priced model.
- vs DeepSeek V4 Pro 0423: DeepSeek V4 Pro costs $0.9553 per 1M input tokens at list, 3.1× cheaper than Claude Sonnet 4.6. It is open-weight, so Guests can call it with no membership fee. Test it on your cost-sensitive work; keep Claude Sonnet 4.6 where your evals favour it.
Related models
- Claude Haiku 4.5
Anthropic's Haiku tier
$1.00 input / $5.00 output at list, 3× cheaper than Claude Sonnet 4.6 at a 3:1 mix; 200K-token context; closed, members only.
- Claude Opus 5.5
Anthropic's newer Opus
$4.00 input / $20.00 output at list, 33% more than Claude Sonnet 4.6 at a 3:1 mix; 1M-token context; closed, members only.
- DeepSeek V4 Pro 0423
DeepSeek's Pro tier
$0.9553 input / $1.91 output at list, 5× cheaper than Claude Sonnet 4.6 at a 3:1 mix; 1M-token context; open-weight, on the Guest plan.
Data policy for this model
Private: Your prompts and outputs are never used for training and never shared or licensed. For security, abuse prevention and debugging we store request and response content, encrypted, and delete it after 30 days; error details may be kept longer. We also keep usage details (model, tokens, cost, latency, status and time). The model provider may keep requests briefly for abuse monitoring under its own terms.
Research Share: available. Opt in per key and pay less. Data from that key will be used to accelerate AGI and supplied to top frontier labs for AGI purposes, and only for traffic after you opt in.
Regions
Each model is served only in the regions its provider supports. Per-model region details are not published yet.
Use Claude Sonnet 4.6 with the OpenAI SDK
Change two settings: point the SDK at https://api.bearer.club/v1 and use your Bearer Club key, then pass anthropic/claude-sonnet-4-6 as the model.
curl https://api.bearer.club/v1/chat/completions \ -H "Authorization: Bearer $BEARER_CLUB_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "model": "anthropic/claude-sonnet-4-6", "messages": [{"role": "user", "content": "Hello from the Club!"}] }'import osfrom openai import OpenAIclient = OpenAI( base_url="https://api.bearer.club/v1", api_key=os.environ["BEARER_CLUB_API_KEY"], # your membership card)resp = client.chat.completions.create( model="anthropic/claude-sonnet-4-6", messages=[{"role": "user", "content": "Hello from the Club!"}],)print(resp.choices[0].message.content)import OpenAI from "openai";const client = new OpenAI({ baseURL: "https://api.bearer.club/v1", apiKey: process.env.BEARER_CLUB_API_KEY, // your membership card});const resp = await client.chat.completions.create({ model: "anthropic/claude-sonnet-4-6", messages: [{ role: "user", content: "Hello from the Club!" }],});console.log(resp.choices[0].message.content);The official OpenAI SDKs and most OpenAI-compatible tools, such as Cursor, Cline, Continue, Open WebUI and LiteLLM, work the same way; the name of the base URL setting varies by tool.
Claude Sonnet 4.6 FAQ
Is Claude Sonnet 4.6 available on the Guest tier?
No. Claude Sonnet 4.6 is a closed model from Anthropic, and it needs a Member ($199 a year) or Executive ($1,999 a year) membership.
What context window does Claude Sonnet 4.6 support?
1,000,000 tokens, roughly 750,000 English words. The prompt and the answer together have to fit in it.
Can I use Claude Sonnet 4.6 with LangChain or LlamaIndex?
Usually, yes. Use the framework's OpenAI-compatible chat client, set the base URL to https://api.bearer.club/v1 and the API key to your Bearer Club key, and use the model id anthropic/claude-sonnet-4-6. The name of the base URL setting varies by tool and version.
Does Bearer Club store my prompts when I use Claude Sonnet 4.6?
On Private keys: Your prompts and outputs are never used for training and never shared or licensed. For security, abuse prevention and debugging we store request and response content, encrypted, and delete it after 30 days. Error details may be kept longer. On Research Share keys you pay less, and prompts and outputs are scrubbed for personal data and secrets and will be used to accelerate AGI and supplied to top frontier labs for AGI purposes.
What is the model ID for Claude Sonnet 4.6 on Bearer Club?
Use anthropic/claude-sonnet-4-6 as the model in requests to https://api.bearer.club/v1/chat/completions. The same id appears in the https://api.bearer.club/v1/models list.
Further reading
- Best API for coding: benchmarks and live prices
Best API for coding in 2026, ranked by task type and live token prices. Covers Claude, DeepSeek, Qwen and GPT, all on one key at member prices.
- OpenRouter alternatives ranked by price and model access
OpenRouter alternatives compared by pricing model, model count and real per-token cost: which service saves most at volume and which has the widest catalog.
- OpenAI-compatible API explained: one key for 69 models
OpenAI-compatible API: change base_url and api_key to reach 69 models including GPT-6, Claude, DeepSeek, and Gemini through Bearer Club at member prices.
Discounts are measured against each provider's public list price and vary by model, membership tier and data policy. See Pricing.