Claude Fable 5 API Relay Pricing on Quanta: $4.35 per 1M Tokens

01 · What does Claude Fable 5 cost for a student research group on Quanta?

Research budgets are usually small, specific and easy to lose track of. A group that needs a capable model for literature review, data annotation or experimental writing should be able to see what each request costs before it runs. Quanta is built for teams with heavy, sustained workloads, and it makes that visibility part of the service rather than an optional extra.

Quanta is a pay-as-you-go API relay with a prepaid balance. There is no monthly fee and no minimum spend, so a student research group can start with a modest balance, measure its real usage and decide what to fund next. Every call is metered on real tokens, and the dashboard lists the tokens and the charge for each one.

For Claude Fable 5, the listed price per 1M tokens is $4.35 for input, $21.74 for output, $5.35 for cache writes and $0.3333 for cache reads. These are approximate USD figures, with charges made in CNY. Because input, output, cache writes and cache reads are priced separately, a group can see which part of its workload is driving the bill instead of receiving a single opaque figure at the end of a billing period.

Claude Fable 5 is available through the official Claude Code and Claude Desktop clients, with Claude Code 2.1.285 or newer. That pairing matters for research groups whose members already write prompts and review results in those tools.

02 · How can a small-budget team forecast its Quanta spending?

Start with the token mix your group actually expects, then use the published rates to build a range. Quanta does not require a commitment in advance, so the forecast can be revised as the project develops.

Here are worked examples at a 3:1 input-to-output ratio with no cache hits:

• › Claude Opus 5.5: 10 million tokens a month ≈ $34.58; 100 million ≈ $345.83.
• › Claude Sonnet 5.5: 10 million tokens a month ≈ $25.07; 100 million ≈ $250.69.
• › GPT-6.1 Sol: 10 million tokens a month ≈ $4.10; 100 million ≈ $40.97.
• › Gemini 2.5 Flash: 10 million tokens a month ≈ $6.49; 100 million ≈ $64.93.

These are examples, not quotas or allowances. A group running many short classification passes may produce a different ratio than one generating long summaries, which is why Quanta reports usage per call. The same reasoning applies to caching: repeated context that can be read from cache is charged at the cache read rate rather than the full input rate, and the dashboard shows those categories separately.

For a tighter budget, the smallest top-up is $1.39. Other available amounts are $6.94, $13.89, $27.78, $69.44 and $138.89. New accounts also receive $0.50 in trial credit on sign-up, which is enough to test a workflow and read the first usage entries before committing a larger balance.

Quanta also publishes referral rewards: 19% of a referred account's top-ups and 5% one level further. Rewards can be moved into the balance, or withdrawn from $27.78. For a group whose members bring in other student projects, that can add usable credit without changing the pay-as-you-go structure.

03 · Can Quanta handle the concurrency a research group needs?

Heavy research workloads often mean many parallel jobs: one script processing papers, another checking citations, a third drafting notes. Quanta publishes and enforces the limits that apply to this kind of work, so nobody has to guess whether a queue will form.

• › 120 requests per minute per key.
• › 8 requests in flight per account.
• › 25 MB per request body.
• › 20 live keys per account.
• › Streams have no total-time cap and end only after 300 seconds of upstream silence.

These are published limits on Quanta, not estimates of what a workload might achieve. A group can size its batch jobs against 8 requests in flight per account, and use separate keys when a single key's 120 requests per minute is the practical constraint. That separation of rate and concurrency limits is useful for sustained runs, because the team can plan its scheduling rather than discovering a ceiling mid-experiment.

The absence of a total-time cap on streams also suits long-running generation tasks. The stream continues until the upstream is silent for 300 seconds, which gives long research sessions room to finish without a short wall-clock deadline cutting them off.

04 · Which clients and models should the group set up first?

Quanta supports three protocols through one key and one balance: OpenAI-compatible (/v1), Anthropic native and Gemini native. The client determines the correct pairing, and the rules are specific.

• › Claude models, including Claude Fable 5, work only in the official Claude Code and Claude Desktop clients, using Claude Code 2.1.285 or newer.
• › Gemini models use the native protocol only, through Gemini CLI.
• › Codex, Cursor, Cherry Studio, Cline, OpenCode, OpenAI SDKs and cURL pair with GPT, DeepSeek and the other non-Claude, non-Gemini models available on Quanta.

For a group starting with Claude Fable 5, the setup is therefore straightforward: use Claude Code or Claude Desktop, connect it to Quanta, and choose the model by its identifier, claude-fable-5. If the lab also wants a lighter model for bulk tagging, Claude Haiku 4.5 is listed at $0.4167 per 1M input tokens and $1.97 per 1M output tokens, with cache writes at $0.4861 and cache reads at $0.0333.

For a mixed toolchain, non-Claude and non-Gemini options include GPT-6.1 Sol at $0.2083 input and $1.01 output per 1M tokens, DeepSeek V4 Flash at $0.2917 input and $1.18 output, and Qwen3.7 Plus at $0.1667 input and $0.4861 output. One key and one balance cover all three protocols, so a group does not need separate accounts for different model families.

05 · How do billing, invoices and refunds work on Quanta?

Quanta charges against a prepaid balance, and the dashboard lists the tokens and charge of every call. There is no monthly fee and no minimum spend, which suits a lab that funds work in bursts around deadlines, data collection windows or paper revisions.

If unused balance needs to come back, it can be refunded within 15 days of the top-up. That short, defined window is worth noting before a group adds a large balance, and it makes the prepaid model easier to reason about.

For institutional payments, company bank transfers can be arranged, and invoices and advance quotes are handled through the Contact page or the Messages page of the dashboard. There is no support email address to hunt for; the same two routes cover questions about company accounts and billing paperwork.

One more detail matters for a research group handling unpublished material: the text users send is kept for 30 days for abuse handling and billing disputes, then deleted, and model answers are not stored. Groups can document that policy in their own lab notes.

06 · What is the first step for a student team?

Create an account on Quanta, claim the $0.50 trial credit and run a small, representative task. Check the dashboard afterwards to see how tokens are split across input, output, cache writes and cache reads. That single call tells the team more than any estimate, because the figures come from its own work.

Then choose a top-up that matches the pace of the project. If the group works in Claude Code or Claude Desktop, Claude Fable 5 is priced at $4.35 per 1M input tokens and $21.74 per 1M output tokens, with cache writes at $5.35 and cache reads at $0.3333. If it needs a second toolchain, the same balance serves the OpenAI-compatible and Gemini native protocols, with Gemini reached through Gemini CLI.

Quanta gives a student research group three things a tight budget depends on: costs that are itemised per call, limits that are stated in advance, and a prepaid balance that can start small. Open an account, meter the first experiment, and let the recorded usage guide the next top-up.