This guide uses source checks from Sep 7, 2026. Provider and gateway prices can change; preserve the checked date with every forecast.
Why This Topic Matters Now
The Sep 6 keyword report is not a broad discovery signal. It is a brand and site-query signal from Tier 1 searchers. `aiwave.live` produced 56 impressions with 55 from Tier 1 countries, while `site:aiwave.live` produced 49 Tier 1 impressions and no clicks. The United States led the country table with 339 impressions. That pattern says evaluators are checking whether AIWave has a credible technical path, but the search result needs to guide them into evidence faster.
A good brand-search path should not depend on a single landing page. It should connect the buyer to docs, model catalog, pricing, trust, privacy, and a working request pattern. This guide treats search results like a procurement queue: a visitor arrives with a brand or site query, then needs enough current evidence to decide whether a redacted trial is worth running. The article is built for Tier 1 and Tier 2 developers, finance reviewers, and security reviewers comparing Chinese AI API access through a unified gateway.
Source Facts Checked Today
AIWave /api/pricing checked on Sep 7, 2026 returned success=true, 63 records, pricing_version a42d372ccf0b5dd13ecf71203521f9d2, default group ratio 3, and VIP group ratio 1. The live pricing page checked the same day still shows the current catalog rows and describes the public base rates as USD per 1M text-token units, with default accounts applying the published multiplier to the base values.
The same AIWave pricing page checked on Sep 7, 2026 lists DeepSeek V4 Flash at $0.638 input, $0.0202884 cache-hit input, and $1.914 output per 1M tokens. It lists DeepSeek V4 Pro at $1.914 input, $0.0637362 cache-hit input, and $5.742 output. It also lists GLM-5.1 at $2.10 input, $0.680001 cache-hit input, and $6.60 output, Kimi K3 at $4.50 input, $0.90 cache-hit input, and $22.50 output, and qwen3.6-flash at $0.2678053 input and $1.60683 output.
The Sep 6 keyword report says official and public market pages checked during the Sep 7 run reinforce token billing, cache-hit economics, context-window tradeoffs, route limits, and OpenAI-compatible migration. It also records the current public AIWave capability language: model routes, 9 providers, one USD invoice, OpenAI-compatible routing, dated rate cards, and per-request ledger language. Public copy should stay with those public capability facts rather than private operating metrics.
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.
| Search query | Evaluator intent | Best proof point |
|---|---|---|
| aiwave.live | Is this the right product? | Home, docs, pricing, trust, console entry |
| site:aiwave.live | Which page should I open? | Search titles that point to docs, models, pricing, and field notes |
| aiwaveblog | Where are technical guides? | Blog index plus model-family field notes |
| aiwave api | Does my OpenAI client work? | Chat Completions and model-list request |
| glm 5 api | Is this route available? | Models catalog, route receipt, source-dated price row |
| chat completion stream | Can it stream safely? | Streaming test, usage receipt, retry policy |
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",
)
models = client.models.list()
candidate_routes = [
model.id for model in models.data
if model.id.startswith(("deepseek", "glm", "kimi", "qwen"))
]
print({
"source_checked_at": "2026-09-07",
"candidate_routes": candidate_routes[:12],
})
Make the Search Result a Route
A brand result should expose a route, not just a name. The title should tell the buyer what they can verify: OpenAI-compatible API, model catalog, USD rate card, trust boundary, or field notes. The description should avoid broad claims and point toward an action the evaluator can complete in one sitting. A strong search path lets someone move from the result page to the model list, run a redacted request, and attach current pricing evidence to an internal review.
Separate Public Proof From Internal Metrics
Current AIWave strategy is explicit: public proof uses capability facts, model catalog, dated pricing, ledger language, and support paths. Private business metrics and private workload narratives stay out of public pages. That matters for every new blog post because older content habits can leak into new copy. This article uses public capability language and current price evidence, but it does not expose internal operating statistics or customer-identifying details.
Use Pricing as a Dated Checkpoint
Pricing should be treated as a checkpoint in the evaluation path. The live endpoint and pricing page were checked on Sep 7, 2026, but a buyer should recheck them before a purchase order or recurring forecast. Store the pricing_version, account group, model ID, input row, cache-hit row, output row, and checked date. If the current request uses the default group, preserve the default multiplier instead of assuming the VIP base rate.
Put Docs Before Registration
Some evaluators are not ready to create an account on the first visit. They need to see the request shape, route names, source dates, and trust boundary first. Internal links should therefore put docs, models, pricing, and trust near the top of the article. Registration can remain a clear call to action after the evidence path is visible. That order fits API buyers who need to bring a short evidence packet back to engineering and procurement.
Tie Model Names to Receipts
A model name seen in a pricing table is not enough for production. The evaluator should run a redacted request, capture the response status and usage object, and store the route name exactly as used. That receipt turns `glm-5`, `kimi-k3`, or `deepseek-v4-flash` from a marketing noun into a tested route. The same receipt also helps the team notice if a fallback or migration changed behavior without review.
Internal Links for Brand Queries
This article should strengthen links to Docs, Chat Completions, Models docs, Pricing, Trust, Privacy, and the API quickstart evidence pack. The goal is a dense technical path, not a broad landing narrative.
Procurement Review
A buyer should be able to attach this path to a procurement note. The note should list the source URLs, checked date, rate-card version, account group, tested route, data boundary, support channel, and recheck owner. It should also explain which facts came from AIWave and which came from an official provider page. That separation keeps the review defensible if direct-provider terms or gateway rows change later.
Final Checklist
For the current keyword cluster, the best response is a tighter evaluation path: search result title, canonical page, docs link, model list, rate-card snapshot, redacted request, usage receipt, trust boundary, and registration. That sequence gives Tier 1 and Tier 2 buyers enough proof to continue without overclaiming. Recheck the live endpoint before purchase decisions, and keep public proof limited to capabilities that can be verified from the site.