Healthcare
Ambient Clinical Scribing
Speech recognition plus LLM products: Abridge, Microsoft Dragon Copilot (unverified), Epic AI Charting (unverified), Ambience.
Named Models
GPT-5.2OpenAIReleasednot in the notesRuns onnot in the notesClassLLMOpenAI model that OpenAI for Healthcare (8 Jan 2026) runs on; no numeric HealthBench score is published for it.Commercial use: proprietary terms; Jakarta hosting: unknown; Indonesian support: not statedGPT-5.5OpenAIReleasedApr 2026Runs onnot in the notesClassLLMOpenAI GPT-5.x model, GA in GitHub Copilot.Commercial use: proprietary terms; Jakarta hosting: unknown; Indonesian support: not statedWhere this comes from
Models the research names for this use case. Research, with sources on each model record.
Who Runs It
production, 4
AbridgeCountry: not in the notesOpenAI for Healthcare APIVendor: OpenAIproductionDate8 Jan 2026Named as API builderEvidence: company statementConfidence: lowSource: openai.com
AmbienceCountry: not in the notesOpenAI for Healthcare APIVendor: OpenAIproductionDate8 Jan 2026Named as API builderEvidence: company statementConfidence: lowSource: openai.com
EMC Healthcare (seven hospitals); also named as adopters: Prodia, EKA Hospital, Pondok Indah Group, Tzu Chi Hospital, Asia One Healthcare, Bali International HospitalCountry: IndonesiaIntelliCare AI-powered electronic medical record with generative AI clinical documentation on InterSystems IRIS for Health; local partners Zi.Care, ICS Compute, BitHealthVendor: InterSystemsproductionDateNov 2025Record completeness at EMC from 68 to 98 percent; seven hospitals in about 18 months (company claims)Evidence: company statementConfidence: lowSource: goodnewsfromindonesia.id
Abridge (300+ health systems)Country: not in the notesAbridge ambient scribeVendor: AbridgeproductionDate–300+ health systems, 100M+ conversations a year; Abridge tested GPT-5.5Evidence: company statementConfidence: lowSource: abridge.com
Where this comes from
Documented uses by named organisations. The evidence column says who made the claim.
Candidate Models
SEA-HELM leaderboard, 58 models, SEA average
A general benchmark for this model class, not a test on this use case. SEA average and per-language (incl. Indonesian) scores for open-weight models. Higher is better. Showing 10 of 11.
Qwen3.6-35B-A3BAlibaba (Qwen) maintainer-run73.86Qwen3.5-35B-A3BAlibaba (Qwen) maintainer-run73.55
gemma-3-27b-itGoogle maintainer-run64.54
Gemma-SEA-LION-v4-27B-ITAI Singapore maintainer-run64.35gpt-oss-20bOpenAI maintainer-run60.64
Llama-3.3-70B-InstructMeta maintainer-run60.44Qwen3-8BAlibaba (Qwen) maintainer-run60.33
GLM 4.7 FlashZ.ai (Zhipu) maintainer-run53.76Gemma-3-4B-itGoogle maintainer-run47.84
Nemotron 3 Nano 30B-A3BNVIDIA maintainer-run46.89050100
Nemotron-SEA-LION-v4.8-120B-A12BAI Singapore, LLMReleasedSep 2026open weightsIndonesian statedhas scoreslast 18 monthsNemotron-SEA-LION-v4.8-30B-A3BAI Singapore, LLMReleasedSep 2026open weightsIndonesian statedhas scoreslast 18 monthsQwen-SEA-LION-v4.5-27B-ITAI Singapore, LLMReleasedMay 2026open weightsIndonesian statedhas scoreslast 18 monthsQwen3.6-35B-A3BAlibaba (Qwen), LLMReleasedApr 2026open weightsruns on edgehas scoreslast 18 monthsQwen3-ASR-0.6BAlibaba (Qwen), speech recognitionReleasedJan 2026open weightsIndonesian statedhas scoreslast 18 monthsQwen3-ASR-1.7BAlibaba (Qwen), speech recognitionReleasedJan 2026open weightsIndonesian statedhas scoreslast 18 monthsNemotron 3 SuperNVIDIA, LLMReleased2026open weightsIndonesian statedhas scoreslast 18 monthsNemotron 3.5 Lightning 30B-A3BNVIDIA, LLMReleased2026open weightsIndonesian statedhas scoreslast 18 monthsgpt-oss-120bOpenAI, LLMReleasedAug 2025open weightsIndonesian statedhas scoreslast 18 months
Granite 4.2 8BIBM, LLMReleasedSep 2026open weightshas scoreslast 18 monthsShowing 10 of 480. All models of this class
Where this comes from
Derived, not researched: models in the directory whose class matches a class this use case uses, excluding family records and models shut down or deprecated. Ranked by how many fit facts they meet, then by release date. Each fit fact is read from the model record.
Links
Who Runs ItAbridgeIn production, 8 Jan 2026AmbienceIn production, 8 Jan 2026EMC Healthcare (seven hospitals); also named as adopters: Prodia, EKA Hospital, Pondok Indah Group, Tzu Chi Hospital, Asia One Healthcare, Bali International HospitalIn production, Nov 2025Abridge (300+ health systems)In production
AbridgeIn production, 8 Jan 2026AmbienceIn production, 8 Jan 2026EMC Healthcare (seven hospitals); also named as adopters: Prodia, EKA Hospital, Pondok Indah Group, Tzu Chi Hospital, Asia One Healthcare, Bali International HospitalIn production, Nov 2025Abridge (300+ health systems)In productionDeployments 4
AbridgeOpenAI for Healthcare APIAmbienceOpenAI for Healthcare APIAbridge (300+ health systems)Abridge ambient scribeEMC Healthcare (seven hospitals); also named as adopters: Prodia, EKA Hospital, Pondok Indah Group, Tzu Chi Hospital, Asia One Healthcare, Bali International HospitalIntelliCare AI-powered electronic medical record with generative AI clinical documentation on InterSystems IRIS for Health; local partners Zi.Care, ICS Compute, BitHealth, Indonesia
Sources
- Generative AI for Clinical Conversations | Abridge abridge.com https://www.abridge.com/
- openai.com/index/openai-for-healthcare https://openai.com/index/openai-for-healthcare/
To cite this page: Project AURION, AI Use-Cases Research & Innovation, "Ambient Clinical Scribing", https://aurion.semai.id/usecases/hc-ambient-scribing/, records of October 2026.