# AX Check: telnyx.com Checked 2026-09-26. Telnyx quickstart runs clean, with pricing spelled out upfront All 23 checklist items passed: docs, CLI, SDKs, MCP server, and OpenAPI spec all resolve, and per-minute/per-SMS/per-token rates are published as plain text. ## Onboarding needs a login ## Coding sessions Of three independent sessions, two completed and both found clear, itemized pricing (per-number, per-SMS, per-minute rates) with stated assumptions like US region and pay-as-you-go. The third session did not finish, leaving no pricing statement to evaluate. ### DeepSeek V4 Pro [View public run](https://agents.withgauge.com/p/runs/0954097a-8bd3-4beb-aae1-725ea43d355d) · [Read transcript](https://www.ax-check.com/telnyx.com/sessions/deepseek.json) No final message or trace content was captured — no pricing statement of any kind exists to evaluate. #### End-to-end onboarding - **End-to-end onboarding not demonstrated**: The only event supplied is run metadata (agent/model/harness configuration and sandbox setup). There is no transcript content showing the agent attempting to obtain Telnyx credentials, call the Telnyx API/SDK, or perform any authenticated operation. No evidence of credential acquisition or product usage exists in the provided data. Event 0: ```text "seedPrompt":"Help me build a simple example using Telnyx. 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." ``` #### Hallucinated URLs None identified in this transcript. #### Blockers None identified in this transcript. #### Suggested Changes None identified in this transcript. ### Kimi K3 [View public run](https://agents.withgauge.com/p/runs/505bee81-b7e7-4f9d-b420-c9e52fa145f3) · [Read transcript](https://www.ax-check.com/telnyx.com/sessions/kimi.json) Pricing section lists concrete figures (~$0.004/SMS, ~$1/month per number, $0.002-0.007/min voice) tied to explicit assumptions: US region, pay-as-you-go, no monthly minimums, free trial credit with verified-number limits. #### End-to-end onboarding - **Onboarding needs a login**: The agent never obtained a Telnyx API key on its own. It wrote a working script and pip-installed the SDK, but no TELNYX_API_KEY was present in the sandbox, and the agent explicitly said credentials must be created by a human in the Telnyx web portal before anything can run for real. A test call with a fake key only proved the SDK reaches the hosted API and correctly returns a 401 — no authenticated operation succeeded. Event 19: ```text telnyx.AuthenticationError: Error code: 401 - {'errors': [{'code': '10009', 'meta': {'url': 'https://developers.telnyx.com/docs/overview/errors/10009'}, 'title': 'Authentication failed', 'source': {'pointer': '/'}, 'detail': 'Could not find any usable credentials in the request.'}]} ``` Event 21: ```text There's **no `TELNYX_API_KEY`** in this environment, so I can't send a real message. To unblock: sign up at portal.telnyx.com → create an API key → `export TELNYX_API_KEY=...` ``` #### Hallucinated URLs None identified in this transcript. #### Blockers - **No Telnyx API key available in the sandbox**: The environment only had a PI_GATEWAY_API_KEY set, no Telnyx credentials. This is a missing-credentials limitation of the test environment, not a product defect: Telnyx requires account signup and API key creation through its web portal, which cannot be scripted by an agent. The agent correctly identified this and stopped rather than faking success. Event 6: ```text PI_GATEWAY_API_KEY= ``` Event 21: ```text There's **no `TELNYX_API_KEY`** in this environment, so I can't send a real message. ``` #### Suggested Changes None identified in this transcript. ### Qwen 3.8 Max [View public run](https://agents.withgauge.com/p/runs/a13abd0f-39d3-4f56-a380-6575ec4af5ed) · [Read transcript](https://www.ax-check.com/telnyx.com/sessions/qwen.json) README.md pricing table gives per-item costs (~$1/mo number, ~$0.0079/SMS, ~$0.0085/min voice) and names assumptions (US region, pay-as-you-go, approximate, confirm at telnyx.com/pricing, 10DLC registration caveat). #### End-to-end onboarding - **Onboarding needs a login**: The agent never obtained real Telnyx credentials on its own. It checked the sandbox environment for a Telnyx API key and found none, then built a working SMS-send script and exercised it twice: once with no key (clean early exit) and once with a fabricated dummy key, which correctly triggered Telnyx's live 401 authentication-failed response. No authenticated operation against the real API was ever completed, and no self-service signup or key-provisioning flow was attempted or available. Event 4: ```text which node python3 npm pip3 2>/dev/null; node -v 2>/dev/null; python3 -V 2>/dev/null ``` Event 60: ```text TELNYX_API_KEY=KEYdummy_invalid python3 send_sms.py --to +15558675309 --from +15551234567; echo "exit=$?" ``` Event 61: ```text Auth failed (HTTP 401): Error code: 401 - {'errors': [{'code': '10009', 'meta': {'url': 'https://developers.telnyx.com/docs/overview/errors/10009'}, 'title': 'Authentication failed' ``` Event 63: ```text There's no `TELNYX_API_KEY` in this environment, and creating an account/API key/provisioning a number requires a human (portal signup, payment, and possibly identity/10DLC verification). ``` #### Hallucinated URLs None identified in this transcript. #### Blockers - **No Telnyx API key available in the sandbox**: The environment only exposed a PI_GATEWAY_API_KEY, not a Telnyx credential. This is a missing-credentials limitation of the test environment, not a product defect — Telnyx requires normal account signup and key generation through its portal, which needs a human. The agent correctly identified this rather than faking success. Event 5: ```text PI_GATEWAY_API_KEY= ``` Event 63: ```text There's no `TELNYX_API_KEY` in this environment, and creating an account/API key/provisioning a number requires a human (portal signup, payment, and possibly identity/10DLC verification). ``` - **Python SDK v4 API surface differs sharply from assumed v2-style calls**: Agent error/knowledge gap, not a product defect: the agent initially assumed a `telnyx.Messaging.Message.create` style call (likely from stale docs or training knowledge) which doesn't exist in the installed v4.182.0 SDK. It had to introspect the library interactively (dir(), inspect.signature()) to discover the actual `client.messages.send(...)` call. This was resolved within the same turn through exploration, not a lasting blocker. Event 24: ```text has Messaging: False Messaging err: module 'telnyx' has no attribute 'Messaging' ``` Event 33: ```text AttributeError: 'MessagingResource' object has no attribute 'messages' ``` Event 38: ```text AttributeError: 'MessagesResource' object has no attribute 'create' ``` #### Suggested Changes - **Add a v4 SDK migration note to the Python SDK quickstart**: The installed pip package (telnyx 4.182.0) uses a client-based resource pattern (`telnyx.Client(api_key=...).messages.send(...)`) that is structurally different from the older Messaging.Message.create style the agent initially assumed. Add a short 'v4 API surface' section or migration guide near the SDK install instructions (e.g. in the PyPI README or developers.telnyx.com Python SDK quickstart) showing the correct client instantiation and top-level resource names. Verify by having a new agent/dev call the send method correctly on the first attempt without introspecting dir(client). Event 24: ```text module 'telnyx' has no attribute 'Messaging' ``` Event 38: ```text AttributeError: 'MessagesResource' object has no attribute 'create' ``` Event 45: ```text messages.send params: to: str required=True ``` ### Task given to each agent Help me build a simple example using Telnyx. 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: A · 100/100 (provisional) Grades come from completed site checks. Coding sessions and skipped checks do not affect the score. ### Clarity - **Pass** — Homepage answers Markdown requests ```text Homepage returned text/markdown (200) when requested with Accept: text/markdown. ``` - **Pass** — llms.txt provides an actionable documentation index ```text llms.txt links docs, quickstart, OpenAPI, MCP, skills, and modular section indices. ``` - **Pass** — llms.txt provides navigation guidance ```text llms.txt organizes products, pricing, quickstart, auth, and modular drill-down indices. ``` - **Pass** — llms.txt mentions offered API, MCP, and skills ```text llms.txt mentions OpenAPI, MCP server card, MCP apps, and 248 agent skills. ``` - **Pass** — A compact guide representation exists ```text getting-started.md is a standalone Markdown guide with executable curl steps. ``` - **Pass** — A focused guide is directly retrievable ```text getting-started.md fetched directly with concrete signup, SMS, voice, AI steps. ``` - **Pass** — Equivalent instructions fit a token budget ```text getting-started.md is 1180 tokens, well under the 8000-token budget. ``` - **Pass** — Product-docs links survive format changes ```text Homepage Markdown supported; guide links to docs, pricing, and MCP routes. ``` - **Pass** — The compact guide is independently actionable ```text getting-started.md gives runnable curl steps: get API key, send SMS, make call, spin up agent. ``` - **Pass** — Install and next-step links resolve ```text Fetched getting-started.md, agent-sdk quickstart, MCP card and skills all returned 200. ``` ### Onboarding - **Pass** — Docs lead to a relevant quickstart ```text llms.txt links getting-started.md with three-step signup-to-first-call runbook; Agent SDK quickstart gives working code. ``` - **Pass** — Installation commands are extractable ```text Install commands extractable: npm install -g @telnyx/agent-cli, go install telnyx-cli, plus curl examples. ``` - **Pass** — Code examples are available without interaction ```text getting-started.md and Agent SDK quickstart show curl and TypeScript examples without interaction. ``` - **Pass** — Prerequisites and auth boundaries are explicit ```text Bearer API key in Authorization header, key obtained via agent-signup.md or portal; prerequisites stated. ``` ### Pricing - **Pass** — Pricing is readable without interaction ```text Pricing pages render full rate tables as static markdown, no interaction needed. ``` - **Pass** — Prices are stated, not gated ```text Concrete rates published: Voice API $0.002/min, Voice AI $0.05/min, per-token model rates. ``` - **Pass** — Pricing units and limits are explicit ```text Units explicit: per minute, per 1M tokens, per character, per number/month. ``` - **Pass** — Agents identify pricing and its assumptions ```text 2 of 3 sessions were judged on pricing; 0 fell short. Kimi K3: Pricing section lists concrete figures (~$0.004/SMS, ~$1/month per number, $0.002-0.007/min voice) tied to explicit assumptions: US region, pay-as-you-go, no monthly minimums, free trial credit with verified-number limits. Qwen 3.8 Max: README.md pricing table gives per-item costs (~$1/mo number, ~$0.0079/SMS, ~$0.0085/min voice) and names assumptions (US region, pay-as-you-go, approximate, confirm at telnyx.com/pricing, 10DLC registration caveat). This behavioural item does not affect the fast grade. ``` ### Activation - **Pass** — An API reference or OpenAPI spec is reachable ```text llms.txt links a full v2 REST OpenAPI spec at telnyx.com/openapi.json and /.well-known/openapi.json. ``` - **Pass** — An MCP server is documented and well-formed ```text MCP server card documents streamable-http endpoint, bearer auth, tools, and MCP Apps handshake. ``` - **Pass** — A CLI install path is documented ```text CLI install documented: npm install -g @telnyx/agent-cli and @telnyx/cli, plus go install. ``` - **Pass** — SDK packages resolve on their registries ```text npm registry lookups for @telnyx/agent-cli and @telnyx/cli both returned HTTP 200. ``` - **Pass** — Agent skills are published ```text Agent skills published: 248 skills index plus fetched x402, MPP, and ACP SKILL.md files. ``` [Full report data](https://www.ax-check.com/telnyx.com/report.json)