# Gradium Voice Cloning and Voice Design vs Hume Octave Voice Design and Cloning + MCP (slim)

> Gradium Voice Cloning and Voice Design scores 60.8 (C) to Hume Octave Voice Design and Cloning's 54.6 (C) for voice cloning. Prices, MCP, x402, uptime and agent notes side by side.

- Full: https://www.anchorterminal.com/compare/gradium-voice-cloning-vs-hume-voice-cloning.md (~5,550 tokens) · this version ~3,080 tokens · JSON https://www.anchorterminal.com/compare/gradium-voice-cloning-vs-hume-voice-cloning.json · canonical https://www.anchorterminal.com/compare/gradium-voice-cloning-vs-hume-voice-cloning
- Index: https://www.anchorterminal.com/llms.txt · API: https://www.anchorterminal.com/api/v1/index.json · MCP: https://www.anchorterminal.com/mcp · Updated: 2026-11-07
- Licence: CC BY 4.0, credit Anchor Terminal and link this page (https://www.anchorterminal.com/terms/#data)

Gradium Voice Cloning and Voice Design scores 60.8 (C) on agent readiness against Hume Octave Voice Design and Cloning + MCP's 54.6 (C), and leads in 4 of 7 scored categories. Hume Octave Voice Design and Cloning + MCP leads on agent ergonomics. Both do voice cloning.

- Gradium Voice Cloning and Voice Design: C 60.8, rank #637 of 1486 · https://www.anchorterminal.com/tools/gradium-voice-cloning.min.md
- Hume Octave Voice Design and Cloning + MCP: C 54.6, rank #899 of 1486 · https://www.anchorterminal.com/tools/hume-voice-cloning.min.md
- Gradium Voice Cloning and Voice Design, good for Builders on Gradium's TTS who want instant clones and designed voices in five European languages with EU residency. Ahead on Reliability, 63 against 58; Security & auth, 48 against 23; Maintenance & community, 70 against 45; Transparency & trust, 63 against 57. Also Free to start without a card.
- Hume Octave Voice Design and Cloning + MCP, good for Designing a voice from a description, and for teams already on EVI. Ahead on Agent ergonomics, 75 against 67. Also Runs on your own machine.

| Category | Gradium Voice Cloning and Voice Design | Hume Octave Voice Design and Cloning + MCP |
| --- | --- | --- |
| Reliability (16%) | 63 | 58 |
| Performance (10%) | pending | pending |
| Schema & documentation (13%) | 83 | 87 |
| Agent ergonomics (13%) | 67 | 75 |
| Security & auth (14%) | 48 | 23 |
| Payments & pricing (10%) | 30 | 30 |
| Task success (10%) | pending | pending |
| Maintenance & community (7%) | 70 | 45 |
| Transparency & trust (7%) | 63 | 57 |

| Fact (where they differ) | Gradium Voice Cloning and Voice Design | Hume Octave Voice Design and Cloning + MCP |
| --- | --- | --- |
| Vendor | Gradium SAS | Hume AI |
| Hosted endpoint | `https://api.gradium.ai/api/voices/` | `https://api.hume.ai/v0/tts` |
| Transports | HTTP, websocket | HTTP, stdio |
| Licence | Proprietary service under Gradium's Terms of Service (version of 19 October 2026). The `gradium` Python SDK is on PyPI | MIT (SDKs) |
| Tools exposed | none | 5 |
| Last release | 2026-11-06 | 2026-08-18 |
| Terms last updated | no date given | 2026-10-05 |
| Privacy policy last updated | no date given | 2025-02-25 |
| Terms restrict automated access | not found in the text | yes |
| Terms restrict benchmarking | not found in the text | yes |
| Terms or service can change without notice | not found in the text | yes |
| Arbitration or class-action waiver | not found in the text | yes |
| Popularity | none | 180 stars, 79k npm/wk, 24k PyPI/wk |
| Agent reviews | none | 2/5 (2) |

Why the edge:
- Security & auth, Gradium Voice Cloning and Voice Design +25: Gradium Voice Cloning and Voice Design 48 (Graded for voice cloning as the other listings in this category, with consent and misuse controls in place of the read-only line and training and retention of voice data in place of the prompt-injection line. API keys per organisation in a header, revocable and able to expire, with no scopes found (20 of 30). No consent or verification field and no watermark. The Terms make the client obtain the consent of anyone whose voice is a sample and bar impersonation without explicit authorisation, and the API FAQ says explicit consent is required (6 of 20). The Terms grant Gradium a licence to use Client Data to improve its services and train its models, with no opt-out found, and zero data retention (paid plans) excludes voice samples, so samples are kept and may be used for training (2 of 15). Organisation roles and a member list, with no per-call audit log found (5 of 15). SOC 2 Type II and ISO 27001:2022 audited by A-LIGN, a trust centre and security@gradium.ai. No security.txt (404) and no bug bounty found (15 of 20). 20+6+2+5+15 = 48.); Hume Octave Voice Design and Cloning + MCP 23 (Graded for voice cloning, with consent and misuse controls in place of the read-only line and training and retention of voice data in place of the prompt-injection line, since the API returns audio and IDs rather than third-party text. One account-wide API key and secret pair, regenerated together (10), and 30 minute OAuth client-credential tokens for clients (5), less 10 because the docs show the API key in a WebSocket query string, `?api_key=` (net 5 of 30). Cloning outside Enterprise happens in the Platform behind a legal agreement checkbox, with no verification (5 of 20). API data isn't used for training, but cloning outside Enterprise is a Platform submission, which may be used unless you opt out in settings, and no retention period is published (5 of 15). Usage is tracked in the portal, no per-call log found for TTS (8 of 15). No security.txt, bug bounty, SOC 2 or trust centre found, HIPAA compliance and a DPA on request are claimed (0 of 20). Per the 1 October check, not re-read on 4 November, because Hume's terms of 5 October 2026 (clause 9.4) bar robots and other automated tools from its websites, so every hume.ai host, status.hume.ai included, is paused.)
- Maintenance & community, Gradium Voice Cloning and Voice Design +25: Gradium Voice Cloning and Voice Design 70 (Read as a model, on the lines the other hosted cloning listings use (recency 30, dated releases 10, deprecation notice 10, support 15, SDKs 15, health 10). Release notes dated 6 November 2026 (data residency in Studio) and 3 November (30). Dated entries on 9, 12, 16, 20, 23 and 28 October, among them voice list filters and the enhancement default for clones on 23 October (10). `cfg_coef` was deprecated with requests still accepted, but the clone enhancement default changed on the day it was announced and no notice policy was found (5 of 10). Public release notes, support@gradium.ai and a Discord server, with no evidence read of how they answer (10 of 15). `gradium` on PyPI at 0.6.4 (4 September 2026), with no SDK in a second language (10 of 15). The SDK repository was not cloned (5 of 10). 30+10+5+10+10+5 = 70.); Hume Octave Voice Design and Cloning + MCP 45 (Read as a closing service, since TTS and EVI API access ends on 13 November 2026, with recency from the last change to the listed product's code, and the umbrella rule for the `hume` SDKs. The last release touching TTS code is Python SDK 0.14.1 (17 August 2026, on PyPI per the 1 October check), built from the Fern regeneration dd9d2c4 of the same day, which sets the version to 0.14.1 and changes the TTS validation error types. It is reached only from a Fern generation tag, not from main or a version tag. 79 days, so 20 of 30, as on 1 October. From 6 August there are only 0.14.1 and the 2 October shutdown notice. TypeScript 0.16.1 (18 August) removed the uuid dependency with no TTS code and isn't counted. That is two dated entries, short of three (0). @humeai/mcp-server's newest tag is 0.3.0 on 8 January 2026. A public changelog, gone quiet, and no answering support channel confirmed (5 of 15). Official Python and TypeScript SDKs (15). TypeScript SDK released within 90 days, Python 3.9 to 3.15 (5 of 10). 20 + 0 + 5 + 15 + 5 = 45.)
- Agent ergonomics, Hume Octave Voice Design and Cloning + MCP +8: Gradium Voice Cloning and Voice Design 67 (Voice objects are compact metadata (20 of 25). `GET /voices/` takes `skip` and `limit` (default 100) and since 23 October filters by `locale`, `gender` and `use_case` (20). Validation errors are structured 422 bodies, while synthesis errors arrive as plain-text HTTP 500 with a numeric code to parse (12 of 20). No idempotency key. Saving a design candidate twice answers 409, and the clone call has none (5 of 20). Three required fields with sensible defaults, and one official SDK, in Python (10 of 15). 20+20+12+5+10 = 67.); Hume Octave Voice Design and Cloning + MCP 75 (The official MCP server has 5 tools, but the JSON TTS endpoint returns base64 audio inside the response, which is heavy in a context window (20 of 25). The voice list paginates and filters by `provider`, so `CUSTOM_VOICE` returns only your voices (20). Named error codes (E0601, E0811 and others) say whether to retry (20). No idempotency key and no job to poll, a design call simply runs again (0 of 20). Few required fields, official Python and TypeScript SDKs (15). Per the 1 October check, not re-read on 4 November, because Hume's terms of 5 October 2026 (clause 9.4) bar robots and other automated tools from its websites, so every hume.ai host, status.hume.ai included, is paused.)
- Transparency & trust, Gradium Voice Cloning and Voice Design +6: Gradium Voice Cloning and Voice Design 63 (Closed service with clear terms, version of 19 October 2026, naming Gradium SAS with its Paris registration (15 of 30). The Terms allow training on Client Data, zero data retention excludes voice samples, the privacy policy gives retention criteria without periods, and a DPA is published as an exhibit. The statements agree but give no retention period for samples (15 of 30). Deprecations appear in dated release notes, with no written notice policy (10 of 20). EU and US residency is documented and selectable from the XS plan. Sub-processors sit in the trust centre, not read (12 of 20). 15+15+10+12 = 52.); Hume Octave Voice Design and Cloning + MCP 57 (Closed service with clear terms, SDKs under MIT (15 of 30). The privacy statement says API data isn't used for training and also that anonymised EVI API data is used by default, a DPA is on request only, and no retention period or subprocessor list is published (10 of 30). Dated deprecations in the changelog, EVI 1 and 2 retired on 30 August 2025 after six weeks' notice (15 of 20). No subprocessor list or data locations found (0 of 20). Per the 1 October check, not re-read on 4 November, because Hume's terms of 5 October 2026 (clause 9.4) bar robots and other automated tools from its websites, so every hume.ai host, status.hume.ai included, is paused.)
- Reliability, Gradium Voice Cloning and Voice Design +5: Gradium Voice Cloning and Voice Design 63 (Graded on the hosted voice API with the hosted lines. Atlassian Statuspage at gradium.statuspage.io with five components, one of them Voice Cloning (20). History and the RSS feed, read on 6 November, show no incident in the 90 days from 8 August. The page has posted one real incident since May 2026 (a DNS issue on 15 June) beside a leftover example incident, so the quiet window is read as minor-only rather than clean, as for tldv (20). Concurrency is published per plan on the pricing page (2 to 15 TTS sessions), while the limits page says to ask support and no limit for voice creation was found (10 of 15). No 429, Retry-After or backoff guidance was found; the errors page says to open a new connection after an error, and voice creation takes a `timeout_s` (3 of 15). No SLA in the Terms or on the pricing page (0). Instant cloning carries no beta label (10). 20+20+10+3+0+10 = 63.); Hume Octave Voice Design and Cloning + MCP 58 (Status page at status.hume.ai (incident.io) with a Text-to-speech API component (20). We couldn't read an incident history on 1 October, the history page returned 404 (5). The sibling EVI listing read the history feed on 1 and 2 October, which holds one TTS incident of about 7 minutes on 19 September; 3 October to 4 November is unread, and a partly unread window keeps its recorded award, so 5 holds. Requests per minute are published by plan on the pricing page, 15 on Free and Starter up to 225 on Business (15). Rate limiting returns error E0811 with 'try again later', and the official SDKs retry with exponential backoff (8, an in-between value kept and flagged). No idempotency or safe-retry guidance for saving a voice (0). No SLA published (0). Voice design over the TTS API is generally available until API access ends on 13 November 2026 (10). 20 + 5 + 15 + 8 + 0 + 0 + 10 = 58. Per the 1 October check, not re-read on 4 November, because Hume's terms of 5 October 2026 (clause 9.4) bar robots and other automated tools from its websites, so every hume.ai host, status.hume.ai included, is paused.)