fone.me vs OpenAI Realtime API

fone.me 23

Closed service. Terms of service exist at /terms but draw by script, so their content, governing law and the company they name are unchecked; no company, address or country of incorporation appears on any page read, llms.txt says San Francisco, and the MCP repository has no licence (5 of 30). Data handling was read only from guide pages, which say only the owner sees their calls and messages, calls are transcribed and callers told so, and private categories are withheld. The privacy policy and data-deletion page draw by script, so no retention period, recording statement, DPA or AI-training statement was read (8 of 30). No deprecation policy or dated notices (0). No subprocessors, telephony, model or hosting providers or data locations are named on any page read; the demo agent card's icon is served from Cloudinary (0 of 20). 5+8+0+0 = 13.

OpenAI Realtime API 90

Closed service with clear terms, per the OpenAI API listing's check of the services agreement, which answered 403 to our reader again on 2 November. The Agents SDK for TypeScript and the OpenAPI file are MIT (15, was 10). The data controls page, re-read on 2 November, lists /v1/realtime with no training use, abuse monitoring logs kept up to 30 days, no stored application state and Zero Data Retention eligibility. The DPA, effective 1 January 2026, commits to return or deletion of customer data on termination, 30 days to object to a new subprocessor, and SCCs with the UK Addendum, and nothing in it conflicts with the guide, per the OpenAI API listing's 4 October check, not re-read on 2 November. The same reading as the other OpenAI listings (30, was 18). The deprecations page, re-read on 2 November, states at least six months' notice for generally available models and lists shutdown dates with replacements, including gpt-realtime and gpt-realtime-mini on 20 January 2027 (20). The guide documents residency for /v1/realtime in the United States and Europe for storage and processing, with a sip-eu.api.openai.com endpoint, and says tracing on /v1/realtime isn't EU residency compliant. The subprocessor list, updated 9 July 2026, names third-party subprocessors and OpenAI affiliates with processing countries for each and a form for notice of new ones, per the OpenAI API listing's 4 October check, not re-read on 2 November. The guide states the residency limit, which counts as disclosure, so the same reading as the other OpenAI listings (20, was 10). 15 + 30 + 20 + 20 = 85.

fone.me 23

Closed service. The newest dated change to the agent surface found is the MCP repository github.com/spearisa/fone-me-mcp, created 5 October 2026 with server.json version 1.0.0, 31 days ago (20). The comparison page marked 'Checked October 28, 2026' and the profile lastmods in the sitemap are content, not releases. No changelog, release notes or three dated changes in 90 days (0 of 20). The support page and FAQ draw by script and no support email, forum or community channel was found; responsiveness unchecked (3 of 15). server.json names the server me.fone.www/agents, a namespace that needs proof of the domain, but the official MCP registry could not be read, so presence is scored absent (0 of 15). No package, public code or CI (0 of 10). 20+0+3+0+0 = 23.

OpenAI Realtime API 82

The last dated changelog entry tagged v1/realtime is the 26 August 2026 deprecation of whisper-1, gpt-4o-transcribe and two related models, which Realtime sessions can use for input transcripts, after gpt-transcribe for committed Realtime turns on 28 July. The listing's own package, @openai/agents-realtime, released 0.21.0 on 20 October 2026, 13 days before the check, with Realtime fixes (WebRTC connections rejected when the server reports an error before the session is configured, Twilio audio defaults kept), read from a clone. A release that changes the product's own package sets the recency line in this run (30, was 20 when SDK releases weren't read on this line). In the 90 days from 4 August the changelog has one v1/realtime entry and the SDK has 11 version tags, 0.14.3 on 5 August to 0.21.0, at least seven of which change the Realtime package's source. SDK tags count half, so 1 + 3.5 clears three (20, was 0). Closed service with a changelog updated several times a week, plus a community forum and Discord linked from the docs, which we did not test (10 of 15). Current official SDKs, @openai/agents and @openai/agents-realtime 0.21.0 (15). The SDK is MIT but still 0.x, and its CI results were not read (7 of 10). 30 + 20 + 10 + 15 + 7 = 82.

fone.me 31

MCP reading, since the MCP server is what an outside agent calls. No tool arguments or input schemas are published in llms.txt, the guide pages or the GitHub README, and robots.txt disallows /mcp, so no session was opened and tools/list was not read. Per-link A2A agent cards are machine-readable JSON (protocol 0.3.0, JSONRPC, input and output modes, skills with examples) and server.json names the remote, but neither types the MCP tools (8 of 25). llms.txt is a real document for agents covering the MCP server, A2A, the search API and the widget (10). Each tool has a one-line purpose in llms.txt and the README, and the how-it-works page states what the agent will and won't do; nothing says when not to use a tool (6 of 20). No typed inputs, enums, constraints or required fields published (0 of 15). Example prompts in the README, per-skill examples in agent cards and widget code for eight site builders; no error response is documented (4 of 15). server.json and the agent cards carry version 1.0.0 and the cards a protocol version; no changelog (3 of 15). 8+10+6+0+4+3 = 31.

OpenAI Realtime API 84

The OpenAPI 3.1 file in openai/openai-openapi (version 2.3.0, last committed 31 October 2026, read from a clone on 2 November) has nine /realtime paths and 185 Realtime schemas, with 11 client events and 46 server events in its main event unions and separate translation session events. It is not an AsyncAPI document, so the WebSocket channel itself is not described (22 of 25). llms.txt and a Markdown twin of every docs page, re-read on 2 November (10). The tools guide has a table of when to use function, mcp with server_url or a connector, and the transport guides say when to choose WebRTC, WebSocket or SIP, per the 9 October check (16 of 20). Session fields carry enums and ranges, such as expires_after.seconds from 10 to 7200, though model also accepts any string (13 of 15). Examples in JavaScript, Python, Java, Ruby and C#, an error event with type, code, message and param, and a list of common MCP failures. No catalogue of Realtime error codes was found (11 of 15). Dated changelog and dated snapshots for the older models (gpt-realtime-2.1 lists only its alias), less 3 because the OpenAPI file still names gpt-4o-realtime-preview models (12 mentions on 2 November) that the deprecations page says were removed on 7 May 2026, and several guides mix GPT-Live and Realtime text on one page (12 of 15). 22 + 10 + 16 + 13 + 11 + 12 = 84.

fone.me 30

No credential exists for callers. llms.txt gives auth 'none', the README says the server is 'public, like the links themselves', and the two agent cards read declare no security scheme. There is nothing to leak or rotate, but fone.me and an owner can tell callers apart only by the name they give ('Maria via ChatGPT' in the README) and by traffic, and no caller identity or verification is documented (10 of 30). Three of four MCP tools are lookups; message_agent and A2A message/send reach any owner's agent. On the owner's side the vendor says the agent asks before agreeing to anything or spending money, the owner approves what it will say before a call, outbound calls run only 8 am to 9 pm local, pause when reported and stop permanently on request, and the owner can take over any chat. No read-only mode and no confirmation on the caller's side (8 of 20). message_agent returns text generated from what the owner taught the agent, and every agent on the platform reads messages from anyone; the only guidance is the how-it-works statement that the agent keeps relationship status, health, money and address private, with no documented prompt-injection or spam handling for the open channel, and the call-screening page says it does not promise to block calls (4 of 15). Owners see every call and chat with a transcript in one inbox; nothing is logged for the calling agent (8 of 15). /.well-known/security.txt answers with the app shell, and no disclosure policy, bug bounty or certification appears on any page read (0 of 20). 10+8+4+8+0 = 30.

OpenAI Realtime API 83

Server connections send the API key as a Bearer header. Browsers use client secrets from POST /v1/realtime/client_secrets that last 10 seconds to 2 hours, 10 minutes by default, and can open several sessions until expiry, and the browser WebSocket example passes the secret as a subprotocol, not in the query string. Project keys take All, Restricted or Read Only permissions, with service accounts, workload identity federation and mutual TLS, per the OpenAI API listing's check. The permissions guide, read on 2 November, puts /v1/realtime under one Model capabilities permission with chat completions, audio, embeddings, images, moderations and responses, so a restricted key can't be limited to Realtime alone (27 of 30, as the OpenAI Speech-to-text listing, was 24 when key scopes were unread). MCP tools take allowed_tools and require_approval, which defaults to always in the OpenAPI file, and a sideband connection keeps tool execution on the server. Session settings attached to a client secret can be overridden by the client connection (15 of 20). The model hears callers and reads MCP results. The SIP guide says to treat SIP headers as untrusted, and no prompt-injection guidance specific to Realtime was found, per the 9 October check (6 of 15). The Admin APIs include audit log retrieval, and since 4 August 2026 the Usage and Costs dashboards and APIs filter and group by API key, the per-key view the other OpenAI listings get 15 for. The Usage API has no Realtime-specific endpoint (15, was 10). openai.com's security.txt, read on 2 November, is PGP-signed with a Bugcrowd programme, a disclosure email and a coordinated disclosure policy, and has no Expires field. trust.openai.com, read on 2 November, names SOC 2 Type 2, ISO/IEC 27001, 27017, 27018, 27701 and 42001 and CSA STAR (20, was 6 when openai.com refused our reader, the award the other OpenAI listings carry). 27 + 15 + 6 + 15 + 20 = 83.

fone.me 15

Hosted reading, for the MCP server and A2A endpoints and the service behind them. No status page is linked from llms.txt, the home page or any guide page read, and https://www.fone.me/status failed to load (0). No readable incident history, so the floor (5). No request rate limit for the MCP server, the search API or A2A is published; the only limits are on the agent's own outbound calls (8 am to 9 pm in the callee's time zone, 'limited per day' with no number), which are not request limits (0 of 15). No 429, overload or retry guidance and no idempotency key for message_agent or message/send (0 of 15). No SLA (0). Nothing on the MCP server, the agent cards or llms.txt is marked beta or preview, and the MCP repository's server.json is version 1.0.0 (10). 0+5+0+0+0+10 = 15.

OpenAI Realtime API 62

Graded as a hosted API. status.openai.com (incident.io) lists Realtime as a component of the API group with history (20). The feed, read on 2 November (102 items, 4 August to 30 October), names Realtime in three incidents in the 90 days from 4 August, and each incident page gives the component's impact. Elevated error rates across API models on 17 September, with Realtime degraded from 20:20 to 21:20 UTC alongside ten other API components, and elevated errors across ChatGPT, Codex and the API on 29 September, with Realtime degraded from 17:52 to 23:14 UTC, are each an hour or more of wide errors, so two majors. Degraded performance across ChatGPT, Codex and the APIs on 21 October listed Realtime for 22 minutes (02:26 to 02:48 UTC), so minor. The 11 August incident on the API, Codex and Work Mode lists no components and isn't counted, as on the other OpenAI listings, and the July incidents the 9 October check counted have left the window. Two majors give 5, as the OpenAI API, Moderation, Embeddings and Speech-to-text listings read the same two incidents (5 of 30, was 20 when the durations were unread). The gpt-realtime-2.1 model page, read on 2 November, gives 400 requests and 200,000 tokens a minute on Build, 10,000 and 4 million on Launch and 20,000 and 15 million on Grow (15, was 5 when the limits were known only from account settings). The docs describe Retry-After on 429 and 503, exponential backoff with jitter and SDK retries, sessions emit rate_limits.updated, and SIP webhooks carry a webhook-id for deduplication, per the 9 October check. No idempotency key for call control was found, so safe-retry guidance without keys (12 of 15). No SLA covering Realtime was found, read as on the other OpenAI listings, where the Scale Tier SLA names GPT and o-series models (0). Generally available since 28 August 2025 (10). 20 + 5 + 15 + 12 + 0 + 10 = 62.

fone.me 29

MCP reading. Four tools named in llms.txt, a small set, but their definitions were not read, so short of full marks (18 of 25). find_agents searches by what someone does and the search API takes a query; no limit, paging or output-size controls are documented (3 of 20). No error codes or messages documented for the MCP server, the search API or A2A (0 of 20). No readOnlyHint or destructiveHint seen and no idempotency key; the three lookups are read-only by name and description, while message_agent writes to a real person's inbox and continues the conversation across calls (2 of 20). No credentials, no configuration beyond the URL, and few stated inputs; no SDK in any language (6 of 15). 18+3+0+2+6 = 29.

OpenAI Realtime API 76

Read on the API lines, per the 9 October check of the guides, not re-read on 2 November. Context can be capped with truncation, retention_ratio and token_limits.post_instructions, items can be deleted, prompt caching is automatic, and response.done reports token usage by modality (22 of 25). There are no lists to page. Output controls are output_modalities, max_output_tokens, optional input transcription and allowed_tools (14 of 20). The error event returns the event_id of the client event that caused it, and MCP tool listing and calls have their own failure events (15 of 20). Client events take an event_id, webhooks carry a webhook-id, and interruptions cancel the response. No session resumption after a dropped connection and no idempotency key for call control were found (10 of 20). A session needs only a model, voice activity detection is automatic, and the Agents SDK starts a browser voice agent in about ten lines, with examples in five languages (15). 22 + 14 + 15 + 10 + 15 = 76.