This guide uses source checks from Sep 11, 2026. Provider and gateway prices can change; preserve the checked date with every forecast.
Why This Topic Matters Now
The Sep 10 keyword report had one genuinely new technical signal: a Tier 1 search for the Moonshot documentation path and `moonshot-v1-128k`. That is small volume, but it is a clean evaluator query. Someone is not asking for brand copy; they are checking a model name, a documentation source, and a long-context route.
A useful Moonshot v1 128K article should therefore behave like a receipt checklist. It should show the current AIWave row, the checked date, the model string, the route owner, the context budget, the output cap, and the evidence packet a procurement reviewer can keep. It should also avoid blending direct Moonshot documentation with AIWave gateway pricing. They are related sources, not interchangeable rows.
Source Facts Checked Today
AIWave /api/pricing checked on Sep 11, 2026 returned success=true, 64 records, pricing_version a42d372ccf0b5dd13ecf71203521f9d2, auto_groups=['default'], and public rows enabled for default, vip, and svip groups. Computed base examples per 1M text-token units were DeepSeek V4 Flash at $0.638 input, $0.020288 cache-hit input, and $1.914 output; DeepSeek V4 Pro at $1.914 input, $0.063736 cache-hit input, and $5.742 output; DeepSeek V3.2 at $0.154 input and $0.308 output; Moonshot v1 128K at $1.80 input and $4.50 output; Moonshot v1 32K at $0.95 input and $2.85 output; Qwen3 Max at $1.5622 input and $6.24879 output; Qwen3.8 2.4T A95B at $2.678053 input and $8.03416 output; and Qwen3.7 Flash at $0.267805 input and $1.071221 output. Actual invoices still depend on account group, final route, request usage, and the receipt captured for that run.
The Sep 10 keyword report says official Moonshot/Kimi pages should be treated as dynamic source material and rechecked before exact provider prices are copied into forecasts. This page uses AIWave live route rows for exact examples and links to Moonshot documentation for direct-provider context.
AIWave current-state decisions keep public content narrow: dated prices, route receipts, machine-readable evidence, trust boundaries, and current probes are acceptable. Private user, revenue, call-volume, paying-customer, customer-workload scale, unsupported uptime, and price-war claims are not used.
Planning Matrix
A source-dated planning matrix keeps the page useful for engineers and procurement reviewers. It turns a search query into an auditable route decision instead of a loose model preference.
| Receipt field | Moonshot reason | Acceptance rule |
|---|---|---|
| model | Route identity must be explicit | Use moonshot-v1-128k after availability check |
| checked_at | Pricing and docs can change | Store Sep 11, 2026 or newer |
| pricing_version | Gateway table needs traceability | a42d372ccf0b5dd13ecf71203521f9d2 |
| context_budget_tokens | 128K context can still be overfilled | Set a per-task ceiling |
| output_cap_tokens | Long answers change spend | Pin by task class |
| source_owner | Direct and gateway rows differ | Separate Moonshot and AIWave fields |
| receipt_id | Review needs evidence | Attach to every accepted trial |
Implementation Pattern
The implementation pattern keeps credentials as placeholders, pins the AIWave base URL, records the model, and leaves room for route-specific controls. Production applications should move credentials into environment or secret storage.
from openai import OpenAI
client = OpenAI(api_key="YOUR_API_KEY_HERE", base_url="https://aiwave.live/v1")
def run_moonshot_receipt(case_id: str, context: str, question: str):
response = client.chat.completions.create(
model="moonshot-v1-128k",
messages=[
{"role": "system", "content": "Answer with risks, assumptions, and next checks."},
{"role": "user", "content": f"Case {case_id}\nContext:\n{context}\nQuestion:\n{question}"},
],
temperature=0.1,
max_tokens=900,
)
return {
"api_key": "YOUR_API_KEY_HERE",
"model": "moonshot-v1-128k",
"source_checked_at": "2026-09-11",
"pricing_version": "a42d372ccf0b5dd13ecf71203521f9d2",
"usage": response.usage,
}
Start With a Long-Context Fit Test
A 128K route is not automatically the right route for every task. Start with three redacted cases: a policy summary, a repository architecture review, and a multi-document support answer. Store the context size, expected answer shape, rejection criteria, and route name before running the model. That makes the long-context decision auditable instead of emotional.
Keep Context Budgets Visible
Context budget should be a configuration value, not a surprise inside the prompt. For each task class, set a maximum input size and a separate output cap. If a case needs more context, record why and rerun with a new receipt. This protects both engineering review and finance review when a long-context route becomes attractive.
Separate Direct Docs From Gateway Rows
Moonshot documentation explains direct-provider behavior and route naming context. AIWave /api/pricing explains live gateway rows for the AIWave account path. Store source_owner, source_url, checked_at, currency, token unit, and route owner separately. A buyer should be able to tell which row came from which surface without asking the engineer who ran the trial.
Record Output Caps by Use Case
A legal-style summary, code review, and support answer should not share one output budget. Pin max_tokens by use case and store it with the receipt. If the model gives a stronger answer only by writing far more tokens, that may be a good tradeoff, but it needs explicit acceptance.
Use Current Probes Carefully
AIWave status is useful as a current probe, not as a historical availability guarantee. During a Moonshot route trial, record the status URL, checked date, request timestamp, and response status. Teams that require availability history should run their own monitoring window or request dated operational evidence.
Internal Links for Moonshot Evaluators
A Moonshot evaluator should move through Models docs, Chat Completions, Pricing, Pricing JSON, Status, Trust, and the docs and pricing sitelink guide. This page adds the Moonshot-specific receipt path.
Procurement Review
Procurement should ask for the model string, live AIWave row, direct Moonshot source link, checked date, pricing_version, context budget, output cap, account group, and a redacted receipt. If any field is missing, the route may still be useful, but the forecast is not ready.
Final Checklist
A Moonshot v1 128K evaluation is ready when route availability is checked, context budgets are configured, output caps are pinned, source owners are separated, receipts include usage fields, and current status is captured without turning one probe into a broad claim. Recheck the live pricing table before procurement.