{"ok":true,"engine":"ssftalk-v1.3-hamiltonian","architecture":"jesse-hamiltonian-v1","lane":"semi_paid_freemium_plus_general_credits","sales_sheet_url":"https://jesse.my/ssftalk","elevenlabs_sheet_url":"https://jesse.my/elevenlabs","checkout_endpoint":"https://jesse.my/api/billing/checkout","verify_session_endpoint":"https://jesse.my/api/billing/verify-session","subscribe_url":"https://jesse.my/account?subscribe=1","credits_store_url":"https://solidsf.com/compute-credits","billing_model":{"free_sandbox_tier":{"daily_free_characters":0,"daily_free_chars":0,"max_chars_per_request":0,"free_studio_preview_seconds":0,"api_key_required":true,"daily_pass_usd":1,"daily_pass_microusd":1000000,"included_endpoints":[]},"paid_credits_tier":{"auth_headers_accepted":["xi-api-key","Authorization: Bearer <key>","x-api-key"],"welcome_grant_microusd":0,"welcome_grant_usd":0,"daily_pass_microusd":1000000,"daily_pass_usd":1,"min_topup_cents":100,"min_topup_usd":1,"unit_scale":"1 USD = 1,000,000 microusd","max_chars_per_request":40000,"store_url":"https://solidsf.com/compute-credits","account_url":"https://jesse.my/account","subscribe_url":"https://jesse.my/account?subscribe=1","sales_sheet_url":"https://jesse.my/ssftalk","checkout_endpoint":"https://jesse.my/api/billing/checkout","verify_session_endpoint":"https://jesse.my/api/billing/verify-session","websocket_url":"wss://api.solidsf.com/v1/ssftalk/ws"},"developer_plan":{"id":"developer_monthly_20","daily_plan_id":"developer_daily_1","name":"Jesse Developer & Voice Pro","price_usd_daily":1,"price_cents_daily":100,"price_usd_monthly":20,"price_cents_monthly":2000,"billing_provider":"stripe","included_api_keys":2,"extra_key_pack_usd_monthly":10,"extra_key_pack_size":2,"welcome_credits_microusd":0,"welcome_credits_usd":0,"daily_pass_microusd":1000000,"daily_pass_usd":1,"checkout_endpoint":"https://jesse.my/api/billing/checkout","checkout_post_body":{"return":"ssftalk"},"verify_session_endpoint":"https://jesse.my/api/billing/verify-session","subscribe_url":"https://jesse.my/account?subscribe=1","account_url":"https://jesse.my/account"}},"semi_paid_policy":{"free_sandbox_tier":{"daily_free_characters":0,"daily_free_chars":0,"max_chars_per_request":0,"free_studio_preview_seconds":0,"api_key_required":true,"daily_pass_usd":1,"daily_pass_microusd":1000000,"included_endpoints":[]},"paid_credits_tier":{"auth_headers_accepted":["xi-api-key","Authorization: Bearer <key>","x-api-key"],"welcome_grant_microusd":0,"welcome_grant_usd":0,"daily_pass_microusd":1000000,"daily_pass_usd":1,"min_topup_cents":100,"min_topup_usd":1,"unit_scale":"1 USD = 1,000,000 microusd","max_chars_per_request":40000,"store_url":"https://solidsf.com/compute-credits","account_url":"https://jesse.my/account","subscribe_url":"https://jesse.my/account?subscribe=1","sales_sheet_url":"https://jesse.my/ssftalk","checkout_endpoint":"https://jesse.my/api/billing/checkout","verify_session_endpoint":"https://jesse.my/api/billing/verify-session","websocket_url":"wss://api.solidsf.com/v1/ssftalk/ws"},"developer_plan":{"id":"developer_monthly_20","daily_plan_id":"developer_daily_1","name":"Jesse Developer & Voice Pro","price_usd_daily":1,"price_cents_daily":100,"price_usd_monthly":20,"price_cents_monthly":2000,"billing_provider":"stripe","included_api_keys":2,"extra_key_pack_usd_monthly":10,"extra_key_pack_size":2,"welcome_credits_microusd":0,"welcome_credits_usd":0,"daily_pass_microusd":1000000,"daily_pass_usd":1,"checkout_endpoint":"https://jesse.my/api/billing/checkout","checkout_post_body":{"return":"ssftalk"},"verify_session_endpoint":"https://jesse.my/api/billing/verify-session","subscribe_url":"https://jesse.my/account?subscribe=1","account_url":"https://jesse.my/account"}},"developer_plan":{"id":"developer_monthly_20","daily_plan_id":"developer_daily_1","name":"Jesse Developer & Voice Pro","price_usd_daily":1,"price_cents_daily":100,"price_usd_monthly":20,"price_cents_monthly":2000,"billing_provider":"stripe","included_api_keys":2,"extra_key_pack_usd_monthly":10,"extra_key_pack_size":2,"welcome_credits_microusd":0,"welcome_credits_usd":0,"daily_pass_microusd":1000000,"daily_pass_usd":1,"checkout_endpoint":"https://jesse.my/api/billing/checkout","checkout_post_body":{"return":"ssftalk"},"verify_session_endpoint":"https://jesse.my/api/billing/verify-session","subscribe_url":"https://jesse.my/account?subscribe=1","account_url":"https://jesse.my/account"},"rate_card":[{"id":"voice.tts.articulate_150m","service":"voice","ws_opcode":2,"ws_engine":"Articulate150mNeural","name":"Jesse v1.3 ARTICULATE-150M 24 kHz Hamiltonian TTS + Timestamps","base_microusd":15,"unit":"utterance_per_1k_chars","elevenlabs_equivalent_microusd":240000,"cost_advantage_vs_elevenlabs":"16.0x cheaper ($0.015 / 1M chars vs $0.24 / 1k chars)","free_tier_included":false,"free_daily_chars":0,"daily_pass_usd":1,"hardware_reservation_required":false,"billing_mode":"paid_daily_pass_or_credits"},{"id":"voice.director.hamiltonian","service":"voice","ws_opcode":3,"ws_engine":"HamiltonianProsodyDirector","name":"Hamiltonian-Fourier Prosody Director + 8-Axis Emotion + 50/50 Actor-Critic Jury","base_microusd":5,"unit":"trajectory","elevenlabs_equivalent_microusd":100000,"cost_advantage_vs_elevenlabs":"20.0x cheaper + deterministic cotangent proof receipt","free_tier_included":false,"free_daily_chars":0,"daily_pass_usd":1,"hardware_reservation_required":false,"billing_mode":"paid_daily_pass_or_credits"},{"id":"voice.studio.gb300_batch","service":"voice","ws_opcode":1,"ws_engine":"Gb300AcousticCfmPtx","name":"Zero-Shot Voice Clone, Speech-to-Speech, Schur-Feshbach Audio Isolation, Port-Hamiltonian SFX & WSOLA Dubbing","base_microusd":25,"unit":"studio_operation","elevenlabs_equivalent_microusd":500000,"cost_advantage_vs_elevenlabs":"20.0x cheaper + 305 us zero-shot clone & 0.000 ms WSOLA drift","free_tier_included":false,"free_preview_seconds":0,"daily_pass_usd":1,"hardware_reservation_required":false,"billing_mode":"paid_daily_pass_or_credits"},{"id":"voice.duplex.diarizer_4d","service":"voice","ws_opcode":4,"ws_engine":"DuplexDiarizer4D","name":"Full-Duplex JAUD Room Turn, 0.30 us Atomic Barge-In, Scribe Diarized ASR & Audio-LLM 88% Token Pruner","base_microusd":10,"unit":"duplex_or_mechinterp_turn","elevenlabs_equivalent_microusd":180000,"cost_advantage_vs_elevenlabs":"18.0x cheaper + 88.0% Audio-LLM KV-cache token savings","free_tier_included":false,"free_daily_chars":0,"daily_pass_usd":1,"hardware_reservation_required":false,"billing_mode":"paid_daily_pass_or_credits"}],"benchmark_14_stage":[{"stage":1,"id":"tts_timestamps","name":"ElevenLabs v1 TTS + Character Timestamps","ssftalk_us":40833.61,"baseline_us":420000,"speedup_x":10.3,"baseline_name":"ElevenLabs Cloud TTS v2 REST","detail":"4.56s 24 kHz audio (112x real-time, RTF=0.00895)"},{"stage":2,"id":"ws_stream_ttfb","name":"WebSocket Stream-Input First-Chunk TTFB","ssftalk_us":24473.17,"baseline_us":75000,"speedup_x":3.1,"baseline_name":"ElevenLabs Flash v2.5 WS","detail":"24.5 ms TTFB + Hann partition-of-unity overlap-add"},{"stage":3,"id":"prosody_director","name":"Hamiltonian Prosody & 8-Axis Emotion Director","ssftalk_us":13.97,"baseline_us":14500,"speedup_x":1038.2,"baseline_name":"Cross-Attn Prosody LLM","detail":"D=64 exact Fourier phase coherence 1.000000"},{"stage":4,"id":"zero_shot_clone","name":"Instant Zero-Shot Voice Cloning (/v1/voices/add)","ssftalk_us":305.3,"baseline_us":480000,"speedup_x":1572.2,"baseline_name":"ECAPA-TDNN Speaker Embed","detail":"64-D Glass-Box F0 + 4-formant vocal tract + 256-bit fingerprint"},{"stage":5,"id":"voice_design","name":"Natural-Language Text-to-Voice Design (3 Previews)","ssftalk_us":62687.4,"baseline_us":1850000,"speedup_x":29.5,"baseline_name":"ElevenLabs Voice Design API","detail":"3 style previews (warm / curious / calm) from desc: prompt"},{"stage":6,"id":"speech_to_speech","name":"Prosody-Preserving Speech-to-Speech (/v1/speech-to-speech)","ssftalk_us":5669.11,"baseline_us":620000,"speedup_x":109.4,"baseline_name":"RVC v2 / ElevenLabs STS","detail":"1.0s 24 kHz (0.96 F0 contour correlation)"},{"stage":7,"id":"audio_isolation","name":"Schur-Feshbach 1+7=8 Studio Audio Isolation","ssftalk_us":73.88,"baseline_us":95000,"speedup_x":1285.9,"baseline_name":"Demucs Spectral U-Net","detail":"+21.9 dB SNR improvement (86.2% noise wavelet detail pruned)"},{"stage":8,"id":"sound_generation","name":"Port-Hamiltonian 12-Mode Physical Sound Generation","ssftalk_us":234.46,"baseline_us":920000,"speedup_x":3924,"baseline_name":"AudioLDM2 Diffusion SFX","detail":"Symplectic 12-mode passive resonator + turbulent burst"},{"stage":9,"id":"wsola_dubbing","name":"Pitch-Invariant WSOLA Multi-Speaker Timeline Dubbing","ssftalk_us":23218.17,"baseline_us":840000,"speedup_x":36.2,"baseline_name":"Librosa Phase-Vocoder Dub","detail":"Multi-speaker EN->FR/DE/ES/JA/ZH with 0.000 ms drift"},{"stage":10,"id":"scribe_align","name":"Scribe Diarization + Forced Alignment","ssftalk_us":1.94,"baseline_us":165000,"speedup_x":85170.5,"baseline_name":"WhisperX + PyAnnote","detail":"Sub-2us word & character forced alignment + speaker diarization"},{"stage":11,"id":"j128_codec","name":"j128 128.0 kbps Predictive Voice Codec Roundtrip","ssftalk_us":4282.95,"baseline_us":6400,"speedup_x":1.5,"baseline_name":"FFmpeg libmp3lame","detail":"Exact 16,000 B/s (128.0 kbps, 15->10B, 0.625 ms latency)"},{"stage":12,"id":"duplex_barge_in","name":"Full-Duplex Room + Atomic Human Barge-In","ssftalk_us":0.3,"baseline_us":3200,"speedup_x":10668.1,"baseline_name":"WebRTC SFU","detail":"12-byte JAUD header + heard_chars atomic floor truncation"},{"stage":13,"id":"audio_llm_prune","name":"Audio-LLM KV-Cache Pruner + 1-Bit b1.xor.popc SAE","ssftalk_us":15.36,"baseline_us":18400,"speedup_x":1198.2,"baseline_name":"Dense Audio-LLM","detail":"88.0% KV-cache tokens pruned + 16x256 b1.xor.popc SAE match"},{"stage":14,"id":"glass_box_sig","name":"64-D Glass-Box Auditable Voice Signature","ssftalk_us":159.28,"baseline_us":28000,"speedup_x":175.8,"baseline_name":"WavLM-Large","detail":"64-D L2-normalized human-auditable vector (self-cos=1.000000)"}],"benchmark_15_stage":[{"stage":1,"id":"tts_timestamps","name":"ElevenLabs v1 TTS + Character Timestamps","ssftalk_us":40833.61,"baseline_us":420000,"speedup_x":10.3,"baseline_name":"ElevenLabs Cloud TTS v2 REST","detail":"4.56s 24 kHz audio (112x real-time, RTF=0.00895)"},{"stage":2,"id":"ws_stream_ttfb","name":"WebSocket Stream-Input First-Chunk TTFB","ssftalk_us":24473.17,"baseline_us":75000,"speedup_x":3.1,"baseline_name":"ElevenLabs Flash v2.5 WS","detail":"24.5 ms TTFB + Hann partition-of-unity overlap-add"},{"stage":3,"id":"prosody_director","name":"Hamiltonian Prosody & 8-Axis Emotion Director","ssftalk_us":13.97,"baseline_us":14500,"speedup_x":1038.2,"baseline_name":"Cross-Attn Prosody LLM","detail":"D=64 exact Fourier phase coherence 1.000000"},{"stage":4,"id":"zero_shot_clone","name":"Instant Zero-Shot Voice Cloning (/v1/voices/add)","ssftalk_us":305.3,"baseline_us":480000,"speedup_x":1572.2,"baseline_name":"ECAPA-TDNN Speaker Embed","detail":"64-D Glass-Box F0 + 4-formant vocal tract + 256-bit fingerprint"},{"stage":5,"id":"voice_design","name":"Natural-Language Text-to-Voice Design (3 Previews)","ssftalk_us":62687.4,"baseline_us":1850000,"speedup_x":29.5,"baseline_name":"ElevenLabs Voice Design API","detail":"3 style previews (warm / curious / calm) from desc: prompt"},{"stage":6,"id":"speech_to_speech","name":"Prosody-Preserving Speech-to-Speech (/v1/speech-to-speech)","ssftalk_us":5669.11,"baseline_us":620000,"speedup_x":109.4,"baseline_name":"RVC v2 / ElevenLabs STS","detail":"1.0s 24 kHz (0.96 F0 contour correlation)"},{"stage":7,"id":"audio_isolation","name":"Schur-Feshbach 1+7=8 Studio Audio Isolation","ssftalk_us":73.88,"baseline_us":95000,"speedup_x":1285.9,"baseline_name":"Demucs Spectral U-Net","detail":"+21.9 dB SNR improvement (86.2% noise wavelet detail pruned)"},{"stage":8,"id":"sound_generation","name":"Port-Hamiltonian 12-Mode Physical Sound Generation","ssftalk_us":234.46,"baseline_us":920000,"speedup_x":3924,"baseline_name":"AudioLDM2 Diffusion SFX","detail":"Symplectic 12-mode passive resonator + turbulent burst"},{"stage":9,"id":"wsola_dubbing","name":"Pitch-Invariant WSOLA Multi-Speaker Timeline Dubbing","ssftalk_us":23218.17,"baseline_us":840000,"speedup_x":36.2,"baseline_name":"Librosa Phase-Vocoder Dub","detail":"Multi-speaker EN->FR/DE/ES/JA/ZH with 0.000 ms drift"},{"stage":10,"id":"scribe_align","name":"Scribe Diarization + Forced Alignment","ssftalk_us":1.94,"baseline_us":165000,"speedup_x":85170.5,"baseline_name":"WhisperX + PyAnnote","detail":"Sub-2us word & character forced alignment + speaker diarization"},{"stage":11,"id":"j128_codec","name":"j128 128.0 kbps Predictive Voice Codec Roundtrip","ssftalk_us":4282.95,"baseline_us":6400,"speedup_x":1.5,"baseline_name":"FFmpeg libmp3lame","detail":"Exact 16,000 B/s (128.0 kbps, 15->10B, 0.625 ms latency)"},{"stage":12,"id":"duplex_barge_in","name":"Full-Duplex Room + Atomic Human Barge-In","ssftalk_us":0.3,"baseline_us":3200,"speedup_x":10668.1,"baseline_name":"WebRTC SFU","detail":"12-byte JAUD header + heard_chars atomic floor truncation"},{"stage":13,"id":"audio_llm_prune","name":"Audio-LLM KV-Cache Pruner + 1-Bit b1.xor.popc SAE","ssftalk_us":15.36,"baseline_us":18400,"speedup_x":1198.2,"baseline_name":"Dense Audio-LLM","detail":"88.0% KV-cache tokens pruned + 16x256 b1.xor.popc SAE match"},{"stage":14,"id":"glass_box_sig","name":"64-D Glass-Box Auditable Voice Signature","ssftalk_us":159.28,"baseline_us":28000,"speedup_x":175.8,"baseline_name":"WavLM-Large","detail":"64-D L2-normalized human-auditable vector (self-cos=1.000000)"},{"stage":15,"id":"audio_studio_bgm_lufs_loop","name":"Audio Studio BGM + EBU R128 LUFS + Seamless Loop","ssftalk_us":1442,"baseline_us":1300000,"speedup_x":901.5,"baseline_name":"Stable Audio 3 + FFmpeg Loudnorm","detail":"48 kHz stereo loop (-16.0 LUFS, -1.0 dBTP, 24-bit WAV/OGG/FLAC)"}],"benchmarks_geomean_speedup_vs_elevenlabs":265.5,"benchmarks_15_stage_geomean_speedup_x":280.6,"audio_studio":{"engine":"ssftalk::studio","mcp_instructions":"solidSF ssftalk Audio Studio (Apple Silicon UMA / NVIDIA Tensor-Core + Cloud WebSocket) for BGM, instrument stems, and sound effects.\ngenerate_audio/edit_audio write 44.1 kHz 24-bit stereo WAV + a .json sidecar (prompt, seed, settings) next to each file;\nprocess_audio trims, loops, fades, normalizes (peak or EBU R128 LUFS) and exports 48 kHz / 44.1 kHz WAV, OGG Vorbis, or 24-bit FLAC for game engines and video editors.\n- Pick audio_type: music -> \"TrackType: Music, VocalType: Instrumental, \", instrument -> \"TrackType: Instrument, \", sfx -> \"TrackType: SFX, \", raw -> no prefix.\n- SFX: fill sound_source, sound_action, sound_material, and sound_space.\n- Generated files receive a safety peak ceiling (-1.0 dBFS default: louder peaks are turned down, never up). Use process_audio to normalize to peak or target LUFS (-14 streaming/video, -16 game BGM, -23 broadcast).","tools":["generate_audio","edit_audio","process_audio","unload_models"],"model_tiers":{"medium":{"dit":"medium","decoder":"same-l","label":"medium + same-l","max_seconds":380},"sm-music":{"dit":"sm-music","decoder":"same-s","label":"sm-music + same-s","max_seconds":120},"sm-sfx":{"dit":"sm-sfx","decoder":"same-s","label":"sm-sfx + same-s","max_seconds":120}},"supported_formats":["wav_pcm24","ogg_vorbis","flac_24"],"sample_rates_hz":[44100,48000]}}