+
AI Dictation in Ollama
Talk to Ollama models running locally on your Mac or PC at full sentence speed — Amical formats prompts for Llama, Mistral, Qwen, and any other open-weight model you have pulled down.
Why typing in Ollama is slowing you down
Long prompts feel like work
Local Llama and Mistral models reward detailed prompts, but typing penalises them. Voice removes the friction so the model on your machine finally gets the architectural context it actually has the capacity to use.
Iteration breaks under typing
Iterating with a local model works best as a conversation — refine, retry, redirect. Typing every turn breaks the natural rhythm voice would close: speak the next instruction, see the next answer, repeat without keyboard fatigue.
Privacy is the whole point
You picked Ollama precisely to keep prompts and responses off third-party servers. Routing dictation through cloud transcription is exactly the privacy contradiction voice should not be reintroducing into your local-first AI workflow.
OllamaWhat Amical does in Ollama
Fully local processing available
Amical local-Whisper option processes everything entirely on your machine — the exact privacy posture Ollama users picked Ollama for, extended to the dictation layer. Audio never leaves your device, mirroring the local-first architecture of the model you are talking to.
Long-form prompting at speed
A multi-paragraph prompt with full architectural intent and prior context goes from idea to submitted in under a minute. Local models with bigger context windows finally get the full prompt they can use, instead of the truncated version typing would have forced.
Multilingual prompting
Ollama supports models tuned for many languages. Amical handles language switches mid-prompt automatically without changing modes — Spanish prompt with English code, Mandarin question with English follow-up, all flowing without keyboard juggling.
What you'd dictate in Ollama
“Multi-paragraph prompt to a local Llama or Mistral model with full architectural context”
“Iterative refinement loop refining an answer through three or four conversational turns”
“Long-context document summary or analysis fully on-device with no cloud round-trip”
Experience the future of dictation
Local and Cloud Models
Choose from powerful cloud-based AI models or run everything locally for maximum privacy and control. Switch seamlessly between providers to find the perfect balance of speed, accuracy, and security.
Custom Vocabulary
Customize the AI to recognize your specific terminology, jargon, and proper nouns for industry-specific accuracy and personalized transcription results.
Project PhoenixMcConaughey
CI/CD pipelinek8s
SOC2PostgreSQL
JiraROI metrics
⌘Space
⌥⌘A
CtrlAltSpace
Custom Shortcuts
Create personalized voice commands and shortcuts to streamline your workflow and boost productivity with hands-free operation.
Multi Language Support
Transcribe and dictate in 100+ languages with native-level accuracy. Switch between languages seamlessly or use mixed-language dictation.
Frequently asked questions
Also works great with
Popular roles using Amical
Coming from another dictation tool?
Loved by builders
From San Francisco to Tokyo — what real users are saying about Amical.
ごめん、勝間さん。
ScribesよりAmicalだわ
This is really an amazing app, and I'm super impressed with how smooth it works, and their amazing support. It's just fast and accurate, and it formats the text depending on what application I'm using. This is really my go-to blathering app, and it makes my day so sweet and easy. Read on AppSumo →
【音声入力】
現状Windowsでも使えるおすすめはこの3つ。
・Amical
・Aqua Voice
・Typeless
個人的には慣れもあるけどAqua Voiceが好き。Typelessは整形キツめなのが難点だけどUIUXが良い。Amicalはローカル完結で無料なのがデカい。
音声入力は移り変わりが激しいので年払いはしない方がいいです。
[Free] Amical for voice input is way too convenient. Reviewing the good and bad of Amical — and walking through installation too.
Translated
有益なアプリをありがとうございます☺️
r/LocalLLMu/todokanai_koi
There's also github.com/amicalhq/amical
Switched to it a couple of weeks back, sleek UI and decent feature set. Works offline and on both windows and mac.
Amical - Open-source AI dictation app (@amical_app)
Cossmology Profile: dub.sh/Vk7tPkn
Key People: Haritabh Singh, Naomi Chopra (@NaomiChopra)
#SpeechRecognition #OpenSource #OSS #COSS
ごめん、勝間さん。
ScribesよりAmicalだわ
This is really an amazing app, and I'm super impressed with how smooth it works, and their amazing support. It's just fast and accurate, and it formats the text depending on what application I'm using. This is really my go-to blathering app, and it makes my day so sweet and easy. Read on AppSumo →
【音声入力】
現状Windowsでも使えるおすすめはこの3つ。
・Amical
・Aqua Voice
・Typeless
個人的には慣れもあるけどAqua Voiceが好き。Typelessは整形キツめなのが難点だけどUIUXが良い。Amicalはローカル完結で無料なのがデカい。
音声入力は移り変わりが激しいので年払いはしない方がいいです。
[Free] Amical for voice input is way too convenient. Reviewing the good and bad of Amical — and walking through installation too.
Translated
有益なアプリをありがとうございます☺️
r/LocalLLMu/todokanai_koi
There's also github.com/amicalhq/amical
Switched to it a couple of weeks back, sleek UI and decent feature set. Works offline and on both windows and mac.
Amical - Open-source AI dictation app (@amical_app)
Cossmology Profile: dub.sh/Vk7tPkn
Key People: Haritabh Singh, Naomi Chopra (@NaomiChopra)
#SpeechRecognition #OpenSource #OSS #COSS
ごめん、勝間さん。
ScribesよりAmicalだわ
This is really an amazing app, and I'm super impressed with how smooth it works, and their amazing support. It's just fast and accurate, and it formats the text depending on what application I'm using. This is really my go-to blathering app, and it makes my day so sweet and easy. Read on AppSumo →
【音声入力】
現状Windowsでも使えるおすすめはこの3つ。
・Amical
・Aqua Voice
・Typeless
個人的には慣れもあるけどAqua Voiceが好き。Typelessは整形キツめなのが難点だけどUIUXが良い。Amicalはローカル完結で無料なのがデカい。
音声入力は移り変わりが激しいので年払いはしない方がいいです。
[Free] Amical for voice input is way too convenient. Reviewing the good and bad of Amical — and walking through installation too.
Translated
有益なアプリをありがとうございます☺️
r/LocalLLMu/todokanai_koi
There's also github.com/amicalhq/amical
Switched to it a couple of weeks back, sleek UI and decent feature set. Works offline and on both windows and mac.
Amical - Open-source AI dictation app (@amical_app)
Cossmology Profile: dub.sh/Vk7tPkn
Key People: Haritabh Singh, Naomi Chopra (@NaomiChopra)
#SpeechRecognition #OpenSource #OSS #COSS
ごめん、勝間さん。
ScribesよりAmicalだわ
This is really an amazing app, and I'm super impressed with how smooth it works, and their amazing support. It's just fast and accurate, and it formats the text depending on what application I'm using. This is really my go-to blathering app, and it makes my day so sweet and easy. Read on AppSumo →
【音声入力】
現状Windowsでも使えるおすすめはこの3つ。
・Amical
・Aqua Voice
・Typeless
個人的には慣れもあるけどAqua Voiceが好き。Typelessは整形キツめなのが難点だけどUIUXが良い。Amicalはローカル完結で無料なのがデカい。
音声入力は移り変わりが激しいので年払いはしない方がいいです。
[Free] Amical for voice input is way too convenient. Reviewing the good and bad of Amical — and walking through installation too.
Translated
有益なアプリをありがとうございます☺️
r/LocalLLMu/todokanai_koi
There's also github.com/amicalhq/amical
Switched to it a couple of weeks back, sleek UI and decent feature set. Works offline and on both windows and mac.
Amical - Open-source AI dictation app (@amical_app)
Cossmology Profile: dub.sh/Vk7tPkn
Key People: Haritabh Singh, Naomi Chopra (@NaomiChopra)
#SpeechRecognition #OpenSource #OSS #COSS
Not sure why nobody's talking about it, but there's a free AI voice input app that's incredibly usable. It's called Amical — I use it daily.
Translated
tried amical (free & open source) alternative to whispr flow, aquavoice because they both charged me over $10
and the fact that it lets me run everything locally so nothing leaves my machine is crazyyy🤯
r/macappsu/IAM-rooted
nice comparison. been using amical lately and surprised it's not mentioned more. it's open source, runs locally or with cloud models, and actually formats stuff based on what app you're in. feels way more usable than just whisper or vosk on their own.
I was using VoiceDash, currently I'm using Blip AI, and now I'm testing Amical. So far, Amical is really, really good in recognizing my language and fantastic in recognizing the English language. I'm really impressed in this short period of time of using it. Read on AppSumo →
Amical for voice input is great — runs locally so it's super fast. I was on Aqua Voice but the 0.5s lag between speaking and seeing text was stressful every time. Local model accuracy is slightly lower but well within acceptable.
Translated
🎊 Big news! Amical @amical_app reached 1,000 stars on GitHub! High five! 🙌
Type 10x faster with AI-powered voice dictation
openalternative.co/amical
Not sure why nobody's talking about it, but there's a free AI voice input app that's incredibly usable. It's called Amical — I use it daily.
Translated
tried amical (free & open source) alternative to whispr flow, aquavoice because they both charged me over $10
and the fact that it lets me run everything locally so nothing leaves my machine is crazyyy🤯
r/macappsu/IAM-rooted
nice comparison. been using amical lately and surprised it's not mentioned more. it's open source, runs locally or with cloud models, and actually formats stuff based on what app you're in. feels way more usable than just whisper or vosk on their own.
I was using VoiceDash, currently I'm using Blip AI, and now I'm testing Amical. So far, Amical is really, really good in recognizing my language and fantastic in recognizing the English language. I'm really impressed in this short period of time of using it. Read on AppSumo →
Amical for voice input is great — runs locally so it's super fast. I was on Aqua Voice but the 0.5s lag between speaking and seeing text was stressful every time. Local model accuracy is slightly lower but well within acceptable.
Translated
🎊 Big news! Amical @amical_app reached 1,000 stars on GitHub! High five! 🙌
Type 10x faster with AI-powered voice dictation
openalternative.co/amical
Not sure why nobody's talking about it, but there's a free AI voice input app that's incredibly usable. It's called Amical — I use it daily.
Translated
tried amical (free & open source) alternative to whispr flow, aquavoice because they both charged me over $10
and the fact that it lets me run everything locally so nothing leaves my machine is crazyyy🤯
r/macappsu/IAM-rooted
nice comparison. been using amical lately and surprised it's not mentioned more. it's open source, runs locally or with cloud models, and actually formats stuff based on what app you're in. feels way more usable than just whisper or vosk on their own.
I was using VoiceDash, currently I'm using Blip AI, and now I'm testing Amical. So far, Amical is really, really good in recognizing my language and fantastic in recognizing the English language. I'm really impressed in this short period of time of using it. Read on AppSumo →
Amical for voice input is great — runs locally so it's super fast. I was on Aqua Voice but the 0.5s lag between speaking and seeing text was stressful every time. Local model accuracy is slightly lower but well within acceptable.
Translated
🎊 Big news! Amical @amical_app reached 1,000 stars on GitHub! High five! 🙌
Type 10x faster with AI-powered voice dictation
openalternative.co/amical
Not sure why nobody's talking about it, but there's a free AI voice input app that's incredibly usable. It's called Amical — I use it daily.
Translated
tried amical (free & open source) alternative to whispr flow, aquavoice because they both charged me over $10
and the fact that it lets me run everything locally so nothing leaves my machine is crazyyy🤯
r/macappsu/IAM-rooted
nice comparison. been using amical lately and surprised it's not mentioned more. it's open source, runs locally or with cloud models, and actually formats stuff based on what app you're in. feels way more usable than just whisper or vosk on their own.
I was using VoiceDash, currently I'm using Blip AI, and now I'm testing Amical. So far, Amical is really, really good in recognizing my language and fantastic in recognizing the English language. I'm really impressed in this short period of time of using it. Read on AppSumo →
Amical for voice input is great — runs locally so it's super fast. I was on Aqua Voice but the 0.5s lag between speaking and seeing text was stressful every time. Local model accuracy is slightly lower but well within acceptable.
Translated
🎊 Big news! Amical @amical_app reached 1,000 stars on GitHub! High five! 🙌
Type 10x faster with AI-powered voice dictation
openalternative.co/amical