Grok 4.6 API Relay Pricing for Support Pipelines on Quanta

01 · What does a support-ticket pipeline actually cost?

A support-ticket pipeline is a heavy, sustained workload by nature. Tickets arrive continuously, each one may need classification, retrieval, drafting and review, and the output text is usually longer than the instructions that produced it. That makes cost estimation a matter of tokens rather than seats, and it is where Quanta's line-by-line approach becomes useful.

Quanta is an API relay built for teams with heavy, sustained workloads. Every call is metered on real tokens, and input, output, cache writes and cache reads are priced separately. The dashboard then lists the tokens and the charge for every individual call, so a pipeline owner can see which stage of the workflow is consuming the balance. There is no monthly fee and no minimum spend, so a team that is testing a ticket workflow pays for what it actually sends and receives.

For orientation, the published price for Grok 4.6 on Quanta is $0.6667 per 1M input tokens, $2.01 per 1M output tokens and $0.1667 per 1M cache reads. Gemini 2.5 Flash is listed at $0.2361 input, $1.89 output, $0.2361 cache write and $0.0236 cache read per 1M tokens, while GPT-6.1 Sol is listed at $0.2083 input, $1.01 output and $0.0403 cache read. All quoted prices are approximate USD figures; charges are made in CNY.

02 · How do the monthly worked examples translate into a ticket queue?

The examples below use a 3:1 input to output ratio with no cache hits, and they are examples rather than quotations for a specific pipeline.

• › 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.

Say your team runs a queue that processes 100 million tokens a month through a mixture of these routes. The example figures above give a usable range for planning, and the per-call records in the Quanta dashboard give the precise figure afterwards. That combination matters for a support operation: the estimate can be prepared in advance, and the actual charge is verifiable line by line rather than inferred from a bundled allowance.

Cache reads deserve particular attention in ticket work, because support prompts often repeat the same policy text, product documentation or prior ticket summaries. Cache reads are priced separately from input, so a route that reuses context can be evaluated on its own terms. The same separation applies to cache writes, which are charged as their own category wherever the selected model supports them.

03 · Which models and clients fit a support-ticket workflow?

Quanta speaks three protocols: OpenAI-compatible at /v1, Anthropic native and Gemini native. One key and one balance cover all three, which means a routing layer can send classification to one family and drafting to another without splitting the team across separate accounts.

Client pairing is specific, and it is worth setting up correctly from the start:

• › Claude models work only from the official Claude Code and Claude Desktop clients, with Claude Code 2.1.285 or newer. They do not run in Cursor, Cherry Studio, Cline, OpenCode, Codex, an SDK or cURL.
• › Gemini models use the native protocol only, through Gemini CLI.
• › Codex, Cursor, Cherry Studio, Cline, OpenCode, OpenAI SDKs and cURL pair only with GPT, DeepSeek and the other non-Claude, non-Gemini models.

That leaves a broad practical selection for ticket automation on Quanta. GPT-6.1 Sol at $0.2083 input and $1.01 output per 1M tokens suits high-volume classification and short drafting tasks. Qwen3.7 Plus is listed at $0.1667 input and $0.4861 output, and DeepSeek V4 Flash at $0.2917 input and $1.18 output with $0.006944 cache reads. For heavier synthesis, Gemini 3.1 Pro is listed at $1.47 input and $8.56 output, and Kimi K3 at $1.31 input and $6.54 output. Grok 4.6, at $0.6667 input and $2.01 output with a $0.1667 cache read rate, is a reasonable middle route for responses that need more than a label but not a long-form answer.

04 · Will concurrency and rate limits hold up during a ticket spike?

Support traffic is uneven. A morning backlog, a product incident or a seasonal surge can multiply the number of in-flight requests within minutes, so the published limits are the first thing to design against. Quanta publishes and enforces the following:

• › 120 requests per minute per key
• › 8 requests in flight per account
• › 25 MB per request body
• › 20 live keys per account

These are stated numbers rather than hidden thresholds, which makes capacity planning a straightforward exercise. A team can size its worker pool against 8 concurrent requests per account, distribute traffic across keys up to the 20 live key limit, and budget request volume against 120 requests per minute per key. When a queue needs to burst, the shape of the ceiling is known in advance.

Long-running generations are not cut off by an arbitrary total-time cap. Streams have no total-time cap and only end after 300 seconds of upstream silence. For a pipeline that streams drafted replies into an agent console, that means a slow model response is allowed to continue rather than being terminated mid-answer.

Request size is bounded at 25 MB per body, which is a practical reminder for ticket pipelines that attach large context bundles. Large reference material can be summarised once and reused, and because cache reads are priced separately, repeated context is visible as its own line in the Quanta dashboard.

05 · How does Quanta keep the invoice readable?

Quanta uses a prepaid balance, and every call is metered on real tokens with input, output, cache writes and cache reads priced separately. The dashboard lists the tokens and charge of every call, so a support lead can trace a single ticket to a single charge and then aggregate a stage, a route or a model. That level of granularity is what makes a pay-as-you-go relay workable for a team that has to explain spend to someone else.

Top-ups are available at $1.39, $6.94, $13.89, $27.78, $69.44 and $138.89, so the balance can be topped up in steps that match the size of the workload rather than a single large commitment. Unused balance can be refunded within 15 days of the top-up. Sign-up includes $0.50 in trial credit, which is enough to run a small pipeline sample and read the first set of itemised charges before committing further funds.

For company accounts, payment is made by bank transfer from a company account. Invoices and advance quotes are arranged through the Contact page or the Messages page of the dashboard, so procurement can request the paperwork in the same place the technical team watches usage. Quanta publishes the pricing structure, the limits and the billing rules directly, which keeps the conversation about the workload instead of about hidden terms.

06 · What should a support team do next?

Start by describing the pipeline as a token flow rather than a request count. Estimate input and output volume, note which steps repeat context, and then compare the published per-1M-token rates on Quanta across the models you are considering. The worked examples above give a starting scale for a monthly budget, and the dashboard converts that estimate into an exact, per-call record once traffic begins.

Then check the client pairing rules for the tools your agents already use, size the queue against the published concurrency and rate limits, and fund the balance in steps that suit the trial. If you want company invoices or an advance quote, the Contact page and the dashboard Messages page are the route to arrange them.

Quanta is built for exactly this shape of work: continuous ticket volume, mixed model routes, and a finance function that needs a defensible number. One key, one balance, published limits, token-level metering and invoices arranged on request. Open an account with the $0.50 trial credit, run a representative slice of your queue, and let the itemised charges on Quanta tell you what the pipeline really costs.