# AX Check: speech-to-text.cloud Checked 2026-09-18. Pricing is clear, but developer docs and setup guidance are missing Plan prices and minute limits are stated openly on the pricing page, but the guide index skips docs, API, and setup instructions entirely. ## Onboarding needs a login ## Coding sessions All three independent sessions (DeepSeek V4 Pro, Kimi K3, Qwen 3.8 Max) completed and found the pricing page readable, reporting per-minute or per-plan rates and the 60 min/month free tier, while each flagged that exact rates depend on model type or region. ### DeepSeek V4 Pro [View public run](https://agents.withgauge.com/p/runs/caf9436e-2f24-468e-a63f-cdf966ab5199) · [Read transcript](https://www.ax-check.com/speech-to-text.cloud/sessions/deepseek.json) Final output gives a pricing table with named assumptions (volume tier 0-500K min/mo, model type standard vs dynamic-batch vs medical, free 60 min/mo tier) sourced from a live curl of the pricing page (seq 18-59). #### End-to-end onboarding - **Onboarding needs a login**: The agent never obtained real Google Cloud credentials. It checked the environment and found no GOOGLE_APPLICATION_CREDENTIALS, no gcloud config, and no service account key. It installed the google-cloud-speech SDK and wrote a transcription script, but the only live call attempt against the hosted API failed immediately with a default-credentials error, and the session ended there with the agent explicitly asking the human to supply a service-account key. Event 13: ```text no gcloud config GOOGLE_APPLICATION_CREDENTIALS= ``` Event 91: ```text Failed: Your default credentials were not found. To set up Application Default Credentials, see https://cloud.google.com/docs/authentication/external/set-up-adc for more information. ``` Event 93: ```text This sandbox has **no GCP credentials** and no `gcloud`, so I can't authenticate to the hosted API. To unblock, either give me a service-account key (I can wire it into the env and run the full transcription), or run the three account-side steps yourself (project → enable Speech-to-Text API → export the key). ``` #### Hallucinated URLs None identified in this transcript. #### Blockers - **No GCP credentials available to call the hosted Speech-to-Text API**: The sandbox had no Google Cloud credentials (no GOOGLE_APPLICATION_CREDENTIALS, no gcloud config) and no gcloud CLI installed. This is a missing-credentials limitation of the test environment, not a product defect — authenticating to a cloud API normally requires account-level setup (project, billing, service account) that an agent cannot self-provision. The agent correctly wrote working SDK code but could not execute an authenticated call, ending the session unresolved. Event 13: ```text no gcloud config GOOGLE_APPLICATION_CREDENTIALS= ``` Event 91: ```text Failed: Your default credentials were not found. To set up Application Default Credentials, see https://cloud.google.com/docs/authentication/external/set-up-adc for more information. ``` #### Suggested Changes None identified in this transcript. ### Kimi K3 [View public run](https://agents.withgauge.com/p/runs/98d89860-34fa-4688-8905-8a1101f4f3b9) · [Read transcript](https://www.ax-check.com/speech-to-text.cloud/sessions/kimi.json) Final output gives $0.006/15s (~$1.44/hr) figure explicitly tied to the V1 API and 60 min/month free tier, and flags that V2/Chirp have different rates and prices may have changed. #### End-to-end onboarding - **Onboarding needs a login**: Agent never obtained real Google Cloud Speech-to-Text credentials during the session. It checked the environment for gcloud CLI, service-account JSON, and the google-cloud-speech Python package, found none, and stopped, writing a local example script it never ran. No authenticated API call was made. Event 6: ```text ModuleNotFoundError: No module named 'google.cloud.speech' ``` Event 12: ```text I'm blocked from running a live demo — this environment has no Google Cloud credentials, no gcloud CLI, and no service-account key. ``` Event 18: ```text I couldn't execute the demo here: this environment has no Google Cloud credentials and the API has no unauthenticated mode. ``` #### Hallucinated URLs None identified in this transcript. #### Blockers - **No Google Cloud credentials available in sandbox**: The sandbox environment lacked any GCP credentials, gcloud CLI, or service-account key, and Google Cloud Speech-to-Text has no anonymous/sandbox mode. This is a test-environment limitation (missing credentials), not a product defect — the agent correctly identified it needed human-provided auth and stopped rather than faking success. Event 6: ```text ModuleNotFoundError: No module named 'google.cloud.speech' ``` Event 10: ```text done ``` Event 12: ```text this environment has no Google Cloud credentials, no `gcloud` CLI, and no service-account key. Google Cloud Speech-to-Text has no anonymous or sandbox mode; every API call requires an authenticated GCP project with billing enabled. ``` #### Suggested Changes None identified in this transcript. ### Qwen 3.8 Max [View public run](https://agents.withgauge.com/p/runs/19e0a11e-c050-45c2-bd2b-ddb6721e1ea1) · [Read transcript](https://www.ax-check.com/speech-to-text.cloud/sessions/qwen.json) Final output gives per-minute rates for standard/enhanced/Chirp models and free-tier limit (60 min/month), while flagging assumptions: 'Exact rates vary by region/model — check the official pricing page before committing.' #### End-to-end onboarding - **Onboarding needs a login**: Agent installed the google-cloud-speech SDK and wrote a plausible example script, but never obtained or used real Google Cloud credentials. It explicitly found no service account JSON, no Application Default Credentials, and no gcloud CLI in the sandbox, and stated it could not run the script or provision a GCP project itself. Event 6: ```text PI_GATEWAY_API_KEY=unused /usr/local/bin/python3 /usr/local/bin/pip3 /usr/local/bin/node Command exited with code 1 ``` Event 18: ```text I cannot actually run it here. The sandbox has no Google Cloud credentials (no service account JSON, no ADC, no `gcloud` CLI). Speech-to-Text requires a GCP project with billing and enabled API — none of which I can provision from here. ``` #### Hallucinated URLs None identified in this transcript. #### Blockers - **No Google Cloud credentials available in sandbox**: The environment had no GCP service account key, no Application Default Credentials, and no gcloud CLI installed. This is a missing-credentials/test-environment limitation rather than a product defect: Speech-to-Text legitimately requires a billed GCP project and API key or service account, which is normal for any GCP product and cannot be self-provisioned without human action. Event 6: ```text PI_GATEWAY_API_KEY=unused /usr/local/bin/python3 /usr/local/bin/pip3 /usr/local/bin/node Command exited with code 1 ``` Event 18: ```text The sandbox has no Google Cloud credentials (no service account JSON, no ADC, no `gcloud` CLI). Speech-to-Text requires a GCP project with billing and enabled API — none of which I can provision from here. ``` #### Suggested Changes - **Add a credential-free trial path to the Speech-to-Text quickstart**: The agent could install the SDK and write correct example code but had no way to exercise the API without a billed GCP project, service account, and enabled API, all requiring manual human setup. Consider documenting a scoped API-key-only path for the REST endpoint in the quickstart so a first call can be verified without full gcloud/ADC setup. Verify by having a fresh environment obtain a credential and successfully call recognize end-to-end without manual gcloud steps. Event 18: ```text The sandbox has no Google Cloud credentials (no service account JSON, no ADC, no `gcloud` CLI). Speech-to-Text requires a GCP project with billing and enabled API — none of which I can provision from here. ``` ### Task given to each agent Help me build a simple example using Speech-to-Text Cloud. Tell me how pricing works, and briefly tell me whether this product will be easy for you to manage. Let me know if you get blocked. If this product has no developer workflow you can act on, say so plainly and stop. Stay light: use the hosted product through its SDK or API. Do not start local service stacks or wait for long-running commands; if the quickstart requires either, say so plainly and stop. No product credentials were supplied and no purchases were authorized. ## Score: D · 40/100 (provisional) Grades come from completed site checks. Coding sessions and skipped checks do not affect the score. ### Clarity - **Failed** — Homepage answers Markdown requests ```text Homepage returned text/html for a Markdown Accept header; no Markdown representation offered. ``` - **Failed** — llms.txt provides an actionable documentation index ```text llms.txt exists but links no docs, API reference, or starting guide; only marketing and legal pages. ``` - **Failed** — llms.txt provides navigation guidance ```text llms.txt lists capabilities and pricing but gives no navigation guidance to documentation sections. ``` - **Failed** — llms.txt mentions offered API, MCP, and skills ```text llms.txt never mentions the API, MCP, or skills surfaces offered elsewhere on the site. ``` - **Pass** — A compact guide representation exists ```text llms.txt is a compact Markdown guide covering capabilities, formats, pricing, and usage steps. ``` - **Pass** — A focused guide is directly retrievable ```text llms.txt fetched directly at /llms.txt with HTTP 200 as text/plain Markdown. ``` - **Pass** — Equivalent instructions fit a token budget ```text llms.txt is 971 tokens, well under the 8000-token budget. ``` - **Skipped** — Product-docs links survive format changes ```text Homepage Markdown unsupported, so link preservation across formats cannot be measured. ``` - **Pass** — The compact guide is independently actionable ```text llms.txt gives concrete steps: upload, configure language/format, transcribe, download. ``` - **Skipped** — Install and next-step links resolve ```text No install or next-step links were fetched; only homepage and llms.txt supplied. ``` ### Onboarding - **Skipped** — Docs lead to a relevant quickstart ```text llms.txt describes a web upload workflow; no fetched docs page or quickstart link is present. ``` - **Skipped** — Installation commands are extractable ```text No installation or CLI/SDK setup instructions appear in the fetched llms.txt content. ``` - **Skipped** — Code examples are available without interaction ```text llms.txt lists capabilities and a 4-step upload flow but no code examples. ``` - **Skipped** — Prerequisites and auth boundaries are explicit ```text llms.txt states no sign-up needed but gives no API auth or prerequisites detail. ``` ### Pricing - **Pass** — Pricing is readable without interaction ```text Pricing page lists all plan prices and features directly in static HTML, no interaction needed. ``` - **Pass** — Prices are stated, not gated ```text Prices stated openly: Basic $4.99/mo, Premium $15.99/mo, Business $29/mo, plus monthly tiers. ``` - **Pass** — Pricing units and limits are explicit ```text Units explicit: 900/3000/8400 minutes per month, text tool action counts, USD, tax excluded. ``` - **Pass** — Agents identify pricing and its assumptions ```text 3 of 3 sessions were judged on pricing; 0 fell short. DeepSeek V4 Pro: Final output gives a pricing table with named assumptions (volume tier 0-500K min/mo, model type standard vs dynamic-batch vs medical, free 60 min/mo tier) sourced from a live curl of the pricing page (seq 18-59). Kimi K3: Final output gives $0.006/15s (~$1.44/hr) figure explicitly tied to the V1 API and 60 min/month free tier, and flags that V2/Chirp have different rates and prices may have changed. Qwen 3.8 Max: Final output gives per-minute rates for standard/enhanced/Chirp models and free-tier limit (60 min/month), while flagging assumptions: 'Exact rates vary by region/model — check the official pricing page before committing.' This behavioural item does not affect the fast grade. ``` ### Activation - **Skipped** — An API reference or OpenAPI spec is reachable ```text llms.txt describes a web converter; no API reference or OpenAPI spec was fetched. ``` - **Skipped** — An MCP server is documented and well-formed ```text No MCP server documentation appears in the fetched llms.txt content. ``` - **Skipped** — A CLI install path is documented ```text llms.txt describes only a web upload workflow; no CLI install path documented. ``` - **Skipped** — SDK packages resolve on their registries ```text No SDK or package registry lookup was supplied for this product. ``` - **Skipped** — Agent skills are published ```text No agent skills are mentioned in the fetched llms.txt content. ``` [Full report data](https://www.ax-check.com/speech-to-text.cloud/report.json)