everypixel-cli 0.5.2__tar.gz → 0.6.1__tar.gz
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- {everypixel_cli-0.5.2 → everypixel_cli-0.6.1}/CHANGELOG.md +10 -0
- {everypixel_cli-0.5.2 → everypixel_cli-0.6.1}/PKG-INFO +57 -9
- {everypixel_cli-0.5.2 → everypixel_cli-0.6.1}/README.md +57 -9
- everypixel_cli-0.6.1/src/everypixel_cli/__init__.py +1 -0
- {everypixel_cli-0.5.2 → everypixel_cli-0.6.1}/src/everypixel_cli/application/services.py +196 -42
- {everypixel_cli-0.5.2 → everypixel_cli-0.6.1}/src/everypixel_cli/cli.py +118 -7
- {everypixel_cli-0.5.2 → everypixel_cli-0.6.1}/src/everypixel_cli/mcp_server.py +71 -5
- {everypixel_cli-0.5.2 → everypixel_cli-0.6.1}/src/everypixel_cli/openapi.py +17 -0
- {everypixel_cli-0.5.2 → everypixel_cli-0.6.1}/src/everypixel_cli/resources/openapi.json +2496 -408
- {everypixel_cli-0.5.2 → everypixel_cli-0.6.1}/src/everypixel_cli/schemas.py +410 -211
- everypixel_cli-0.5.2/src/everypixel_cli/__init__.py +0 -1
- {everypixel_cli-0.5.2 → everypixel_cli-0.6.1}/.gitignore +0 -0
- {everypixel_cli-0.5.2 → everypixel_cli-0.6.1}/LICENSE +0 -0
- {everypixel_cli-0.5.2 → everypixel_cli-0.6.1}/pyproject.toml +0 -0
- {everypixel_cli-0.5.2 → everypixel_cli-0.6.1}/src/everypixel_cli/__main__.py +0 -0
- {everypixel_cli-0.5.2 → everypixel_cli-0.6.1}/src/everypixel_cli/application/__init__.py +0 -0
- {everypixel_cli-0.5.2 → everypixel_cli-0.6.1}/src/everypixel_cli/application/chat.py +0 -0
- {everypixel_cli-0.5.2 → everypixel_cli-0.6.1}/src/everypixel_cli/application/models.py +0 -0
- {everypixel_cli-0.5.2 → everypixel_cli-0.6.1}/src/everypixel_cli/application/serialization.py +0 -0
- {everypixel_cli-0.5.2 → everypixel_cli-0.6.1}/src/everypixel_cli/client.py +0 -0
- {everypixel_cli-0.5.2 → everypixel_cli-0.6.1}/src/everypixel_cli/config.py +0 -0
- {everypixel_cli-0.5.2 → everypixel_cli-0.6.1}/src/everypixel_cli/errors.py +0 -0
- {everypixel_cli-0.5.2 → everypixel_cli-0.6.1}/src/everypixel_cli/files.py +0 -0
- {everypixel_cli-0.5.2 → everypixel_cli-0.6.1}/src/everypixel_cli/output.py +0 -0
- {everypixel_cli-0.5.2 → everypixel_cli-0.6.1}/src/everypixel_cli/resources/__init__.py +0 -0
|
@@ -1,5 +1,15 @@
|
|
|
1
1
|
# Changelog
|
|
2
2
|
|
|
3
|
+
## [0.6.1] - 2026-11-18
|
|
4
|
+
|
|
5
|
+
- Add Gemini Omni 1.2 Flash video generation and editing.
|
|
6
|
+
- Add HeyGen Avatar 6 and Turbo photo animation with audio and 2K output.
|
|
7
|
+
|
|
8
|
+
## [0.6.0] - 2026-11-16
|
|
9
|
+
|
|
10
|
+
- Add GPT Image 2.6 Sunburst with format, compression, and animation controls.
|
|
11
|
+
- Add HeyGen Avatar 4, video lipsync, translation, and voice/language catalogs.
|
|
12
|
+
|
|
3
13
|
## [0.5.2] - 2026-11-04
|
|
4
14
|
|
|
5
15
|
- Add Seedance seeds and Seedance 2.5 draft previews with final rendering.
|
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
Metadata-Version: 2.5
|
|
2
2
|
Name: everypixel-cli
|
|
3
|
-
Version: 0.
|
|
3
|
+
Version: 0.6.1
|
|
4
4
|
Summary: Official command line interface for Everypixel API
|
|
5
5
|
Project-URL: Homepage, https://labs.everypixel.com
|
|
6
6
|
Project-URL: Repository, https://github.com/everypixel/everypixel-cli
|
|
@@ -198,15 +198,30 @@ Recraft V4.1 vector models (`recraftv4_1_vector`, `recraftv4_1_pro_vector`)
|
|
|
198
198
|
produce SVG files and accept optional JSON palette `--controls`. Use `--size`
|
|
199
199
|
to select their aspect ratio. Raster models also accept `--resolution`.
|
|
200
200
|
|
|
201
|
-
`--lora-url`
|
|
201
|
+
`--lora-url` applies only to `zimage`. Source-image generation with `--image`
|
|
202
|
+
uses `zimage` or `flux2 --style transparent`; use `image edit` for other modes.
|
|
203
|
+
`--megapixel-ratio` applies only to `flux2` and `qwen` editing; these editing
|
|
204
|
+
models do not accept `--resolution`.
|
|
202
205
|
|
|
203
|
-
GPT Image models are `gpt-image-2
|
|
204
|
-
and editing accept
|
|
205
|
-
|
|
206
|
+
GPT Image models are `gpt-image-2`, `gpt-image-2.5-sunburst`, and
|
|
207
|
+
`gpt-image-2.6-sunburst`. Generation and editing accept
|
|
208
|
+
`--quality low|medium|high` (default: `medium`); both Sunburst models also accept
|
|
209
|
+
`xhigh` and `max`. Use `--size square` with `--resolution 3k`.
|
|
210
|
+
|
|
211
|
+
GPT Image 2.6 accepts `--complexity low|medium|high` and
|
|
212
|
+
`--output-format png|jpeg|webp|gif`; generation also accepts `svg`.
|
|
213
|
+
`--output-compression 0..100` applies only to JPEG/WebP. GIF accepts
|
|
214
|
+
`--frame-count 2..16` and `--frame-duration-ms 20..2000`; omit `--size` and
|
|
215
|
+
`--resolution` because GIF sizing is automatic. These controls apply only to
|
|
216
|
+
GPT Image 2.6.
|
|
206
217
|
|
|
207
218
|
```bash
|
|
208
219
|
everypixel image generate "a detailed botanical illustration" \
|
|
209
|
-
--model gpt-image-2.
|
|
220
|
+
--model gpt-image-2.6-sunburst --quality high --download ./outputs
|
|
221
|
+
|
|
222
|
+
everypixel image generate "a looping animation of a waving cat" \
|
|
223
|
+
--model gpt-image-2.6-sunburst --output-format gif --frame-count 9 \
|
|
224
|
+
--frame-duration-ms 200 --download ./outputs
|
|
210
225
|
```
|
|
211
226
|
|
|
212
227
|
Other image models include `flux3`, `gemini-nano-banana-2.1` (Nano Banana 2.1),
|
|
@@ -234,16 +249,16 @@ everypixel video edit "continue the scene with a slow camera pan" \
|
|
|
234
249
|
--model flux3 --video ./input.mp4 --download ./outputs
|
|
235
250
|
```
|
|
236
251
|
|
|
237
|
-
Generate a video with Gemini Omni 1.
|
|
252
|
+
Generate a video with Gemini Omni 1.2 Flash, then edit the completed task:
|
|
238
253
|
|
|
239
254
|
```bash
|
|
240
255
|
task_id="$(
|
|
241
256
|
everypixel video generate "a slow aerial shot over mountain lakes" \
|
|
242
|
-
--model gemini-omni-1.
|
|
257
|
+
--model gemini-omni-1.2-flash --output-json | jq -er '.task_id'
|
|
243
258
|
)"
|
|
244
259
|
|
|
245
260
|
everypixel video edit "change the scene to sunset" \
|
|
246
|
-
--model gemini-omni-1.
|
|
261
|
+
--model gemini-omni-1.2-flash --task-id "$task_id" --download ./outputs
|
|
247
262
|
```
|
|
248
263
|
|
|
249
264
|
Gemini Omni editing requires an owned Gemini Omni source task. Other editing
|
|
@@ -314,6 +329,39 @@ everypixel audio tts-create --model qwen3 --text "Hello world" \
|
|
|
314
329
|
--speaker Ryan --style Warm --language English --download ./outputs
|
|
315
330
|
```
|
|
316
331
|
|
|
332
|
+
HeyGen video lipsync accepts `--model heygen-speed|heygen-precision` with
|
|
333
|
+
`--video` and `--audio`. Audio and video durations must differ by at most 15%.
|
|
334
|
+
HeyGen Avatar 4 uses `lipsync image --model heygen-avatar-4` with either
|
|
335
|
+
`--audio` or `--text` plus `--voice-id`. Avatar 4 options are
|
|
336
|
+
`--aspect-ratio`, `--expressiveness low|medium|high`, and `--resolution 720p|1080p`
|
|
337
|
+
(default: `1080p`); `--prompt` directs motion. Avatar 4 does not accept `--seed`.
|
|
338
|
+
|
|
339
|
+
HeyGen Avatar 6 (`heygen-avatar-6`, `heygen-avatar-6-turbo`) requires `--image`
|
|
340
|
+
and `--audio`. It accepts
|
|
341
|
+
`--resolution 1080p|2k` (default: `1080p`), preserves the photo's aspect ratio,
|
|
342
|
+
and supports `--expressiveness default|less` plus a motion `--prompt`
|
|
343
|
+
(up to 1200 characters). Text, voice ID, aspect ratio, and seed options belong
|
|
344
|
+
to other models.
|
|
345
|
+
|
|
346
|
+
```bash
|
|
347
|
+
everypixel lipsync heygen-voices --language English --limit 20
|
|
348
|
+
|
|
349
|
+
everypixel lipsync image --model heygen-avatar-4 --image ./portrait.png \
|
|
350
|
+
--text "Hello!" --voice-id VOICE_ID --download ./outputs
|
|
351
|
+
|
|
352
|
+
everypixel lipsync image --model heygen-avatar-6 --image ./portrait.png \
|
|
353
|
+
--audio ./voice.mp3 --resolution 2k --expressiveness less \
|
|
354
|
+
--download ./outputs
|
|
355
|
+
|
|
356
|
+
everypixel video translation-languages
|
|
357
|
+
|
|
358
|
+
everypixel video translate --video ./input.mp4 --output-language Spanish \
|
|
359
|
+
--model heygen-precision --download ./outputs
|
|
360
|
+
```
|
|
361
|
+
|
|
362
|
+
Translation creates one task for one target language. Use `--estimate-cost`
|
|
363
|
+
with HeyGen lipsync or translation to request a quote.
|
|
364
|
+
|
|
317
365
|
`--speaker`, `--style`, `--language`, and `--seed` are Qwen controls;
|
|
318
366
|
`--character` applies only to Qwen voice design. ElevenLabs synthesis uses
|
|
319
367
|
`--voice-id` instead of preset controls. `tts-clone` and `tts-design` default
|
|
@@ -164,15 +164,30 @@ Recraft V4.1 vector models (`recraftv4_1_vector`, `recraftv4_1_pro_vector`)
|
|
|
164
164
|
produce SVG files and accept optional JSON palette `--controls`. Use `--size`
|
|
165
165
|
to select their aspect ratio. Raster models also accept `--resolution`.
|
|
166
166
|
|
|
167
|
-
`--lora-url`
|
|
168
|
-
|
|
169
|
-
|
|
170
|
-
|
|
171
|
-
|
|
167
|
+
`--lora-url` applies only to `zimage`. Source-image generation with `--image`
|
|
168
|
+
uses `zimage` or `flux2 --style transparent`; use `image edit` for other modes.
|
|
169
|
+
`--megapixel-ratio` applies only to `flux2` and `qwen` editing; these editing
|
|
170
|
+
models do not accept `--resolution`.
|
|
171
|
+
|
|
172
|
+
GPT Image models are `gpt-image-2`, `gpt-image-2.5-sunburst`, and
|
|
173
|
+
`gpt-image-2.6-sunburst`. Generation and editing accept
|
|
174
|
+
`--quality low|medium|high` (default: `medium`); both Sunburst models also accept
|
|
175
|
+
`xhigh` and `max`. Use `--size square` with `--resolution 3k`.
|
|
176
|
+
|
|
177
|
+
GPT Image 2.6 accepts `--complexity low|medium|high` and
|
|
178
|
+
`--output-format png|jpeg|webp|gif`; generation also accepts `svg`.
|
|
179
|
+
`--output-compression 0..100` applies only to JPEG/WebP. GIF accepts
|
|
180
|
+
`--frame-count 2..16` and `--frame-duration-ms 20..2000`; omit `--size` and
|
|
181
|
+
`--resolution` because GIF sizing is automatic. These controls apply only to
|
|
182
|
+
GPT Image 2.6.
|
|
172
183
|
|
|
173
184
|
```bash
|
|
174
185
|
everypixel image generate "a detailed botanical illustration" \
|
|
175
|
-
--model gpt-image-2.
|
|
186
|
+
--model gpt-image-2.6-sunburst --quality high --download ./outputs
|
|
187
|
+
|
|
188
|
+
everypixel image generate "a looping animation of a waving cat" \
|
|
189
|
+
--model gpt-image-2.6-sunburst --output-format gif --frame-count 9 \
|
|
190
|
+
--frame-duration-ms 200 --download ./outputs
|
|
176
191
|
```
|
|
177
192
|
|
|
178
193
|
Other image models include `flux3`, `gemini-nano-banana-2.1` (Nano Banana 2.1),
|
|
@@ -200,16 +215,16 @@ everypixel video edit "continue the scene with a slow camera pan" \
|
|
|
200
215
|
--model flux3 --video ./input.mp4 --download ./outputs
|
|
201
216
|
```
|
|
202
217
|
|
|
203
|
-
Generate a video with Gemini Omni 1.
|
|
218
|
+
Generate a video with Gemini Omni 1.2 Flash, then edit the completed task:
|
|
204
219
|
|
|
205
220
|
```bash
|
|
206
221
|
task_id="$(
|
|
207
222
|
everypixel video generate "a slow aerial shot over mountain lakes" \
|
|
208
|
-
--model gemini-omni-1.
|
|
223
|
+
--model gemini-omni-1.2-flash --output-json | jq -er '.task_id'
|
|
209
224
|
)"
|
|
210
225
|
|
|
211
226
|
everypixel video edit "change the scene to sunset" \
|
|
212
|
-
--model gemini-omni-1.
|
|
227
|
+
--model gemini-omni-1.2-flash --task-id "$task_id" --download ./outputs
|
|
213
228
|
```
|
|
214
229
|
|
|
215
230
|
Gemini Omni editing requires an owned Gemini Omni source task. Other editing
|
|
@@ -280,6 +295,39 @@ everypixel audio tts-create --model qwen3 --text "Hello world" \
|
|
|
280
295
|
--speaker Ryan --style Warm --language English --download ./outputs
|
|
281
296
|
```
|
|
282
297
|
|
|
298
|
+
HeyGen video lipsync accepts `--model heygen-speed|heygen-precision` with
|
|
299
|
+
`--video` and `--audio`. Audio and video durations must differ by at most 15%.
|
|
300
|
+
HeyGen Avatar 4 uses `lipsync image --model heygen-avatar-4` with either
|
|
301
|
+
`--audio` or `--text` plus `--voice-id`. Avatar 4 options are
|
|
302
|
+
`--aspect-ratio`, `--expressiveness low|medium|high`, and `--resolution 720p|1080p`
|
|
303
|
+
(default: `1080p`); `--prompt` directs motion. Avatar 4 does not accept `--seed`.
|
|
304
|
+
|
|
305
|
+
HeyGen Avatar 6 (`heygen-avatar-6`, `heygen-avatar-6-turbo`) requires `--image`
|
|
306
|
+
and `--audio`. It accepts
|
|
307
|
+
`--resolution 1080p|2k` (default: `1080p`), preserves the photo's aspect ratio,
|
|
308
|
+
and supports `--expressiveness default|less` plus a motion `--prompt`
|
|
309
|
+
(up to 1200 characters). Text, voice ID, aspect ratio, and seed options belong
|
|
310
|
+
to other models.
|
|
311
|
+
|
|
312
|
+
```bash
|
|
313
|
+
everypixel lipsync heygen-voices --language English --limit 20
|
|
314
|
+
|
|
315
|
+
everypixel lipsync image --model heygen-avatar-4 --image ./portrait.png \
|
|
316
|
+
--text "Hello!" --voice-id VOICE_ID --download ./outputs
|
|
317
|
+
|
|
318
|
+
everypixel lipsync image --model heygen-avatar-6 --image ./portrait.png \
|
|
319
|
+
--audio ./voice.mp3 --resolution 2k --expressiveness less \
|
|
320
|
+
--download ./outputs
|
|
321
|
+
|
|
322
|
+
everypixel video translation-languages
|
|
323
|
+
|
|
324
|
+
everypixel video translate --video ./input.mp4 --output-language Spanish \
|
|
325
|
+
--model heygen-precision --download ./outputs
|
|
326
|
+
```
|
|
327
|
+
|
|
328
|
+
Translation creates one task for one target language. Use `--estimate-cost`
|
|
329
|
+
with HeyGen lipsync or translation to request a quote.
|
|
330
|
+
|
|
283
331
|
`--speaker`, `--style`, `--language`, and `--seed` are Qwen controls;
|
|
284
332
|
`--character` applies only to Qwen voice design. ElevenLabs synthesis uses
|
|
285
333
|
`--voice-id` instead of preset controls. `tts-clone` and `tts-design` default
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
__version__ = "0.6.1"
|
|
@@ -42,14 +42,19 @@ from ..schemas import (
|
|
|
42
42
|
TTSVoiceSavePayload,
|
|
43
43
|
TTSVoicesPayload,
|
|
44
44
|
TTSVoiceDeletePayload,
|
|
45
|
-
|
|
45
|
+
HeyGenImageLipsyncPayload,
|
|
46
|
+
HeyGenAvatar6Payload,
|
|
47
|
+
HeyGenVoicesPayload,
|
|
48
|
+
VideoTranslatePayload,
|
|
49
|
+
RecraftVectorModel,
|
|
50
|
+
ImageEditModel,
|
|
51
|
+
ImageGenerateModel,
|
|
46
52
|
ImageEditPayload,
|
|
47
53
|
ImageGeneratePayload,
|
|
48
54
|
ImageUpscalePayload,
|
|
49
55
|
ImageVectorizePayload,
|
|
50
56
|
LipsyncImagePayload,
|
|
51
57
|
LipsyncVideoPayload,
|
|
52
|
-
RecraftGeneratePayload,
|
|
53
58
|
TaskResponse,
|
|
54
59
|
VideoEditRequest,
|
|
55
60
|
VideoGenerateRequest,
|
|
@@ -79,6 +84,7 @@ COST_ESTIMATE_ENDPOINTS = frozenset(
|
|
|
79
84
|
"image_edit_colors",
|
|
80
85
|
"video_lipsync",
|
|
81
86
|
"image_lipsync",
|
|
87
|
+
"video_translate",
|
|
82
88
|
"video_generate",
|
|
83
89
|
"video_edit",
|
|
84
90
|
"video_upscale",
|
|
@@ -655,6 +661,7 @@ def _build_omni_video_edit_payload(inputs: _VideoEditInputs) -> dict[str, Any]:
|
|
|
655
661
|
|
|
656
662
|
_VIDEO_EDIT_BUILDERS: dict[str, Callable[[_VideoEditInputs], dict[str, Any]]] = {
|
|
657
663
|
"gemini-omni-1.1-flash": _build_omni_video_edit_payload,
|
|
664
|
+
"gemini-omni-1.2-flash": _build_omni_video_edit_payload,
|
|
658
665
|
"seedance2": _build_seedance_video_edit_payload,
|
|
659
666
|
"seedance2-mini": _build_seedance_video_edit_payload,
|
|
660
667
|
"seedance2.5": _build_seedance_video_edit_payload,
|
|
@@ -717,10 +724,26 @@ def build_video_edit_payload(
|
|
|
717
724
|
return builder(inputs)
|
|
718
725
|
|
|
719
726
|
|
|
720
|
-
|
|
721
|
-
|
|
722
|
-
|
|
723
|
-
|
|
727
|
+
_IMAGE_GENERATE_ADAPTER: TypeAdapter[ImageGeneratePayload] = TypeAdapter(
|
|
728
|
+
ImageGeneratePayload
|
|
729
|
+
)
|
|
730
|
+
_IMAGE_EDIT_ADAPTER: TypeAdapter[ImageEditPayload] = TypeAdapter(ImageEditPayload)
|
|
731
|
+
|
|
732
|
+
|
|
733
|
+
def dump_image_payload(payload: BaseModel) -> dict[str, Any]:
|
|
734
|
+
# GIF sizing is automatic; sending materialized size defaults is invalid.
|
|
735
|
+
exclude = (
|
|
736
|
+
{"image_size", "resolution"}
|
|
737
|
+
if getattr(payload, "output_format", None) == "gif"
|
|
738
|
+
else set()
|
|
739
|
+
)
|
|
740
|
+
return payload.model_dump(mode="json", exclude_none=True, exclude=exclude)
|
|
741
|
+
|
|
742
|
+
|
|
743
|
+
def image_fallback_extension(payload: Mapping[str, Any]) -> str:
|
|
744
|
+
if payload.get("model") in get_args(RecraftVectorModel):
|
|
745
|
+
return ".svg"
|
|
746
|
+
return "." + str(payload.get("output_format", "png"))
|
|
724
747
|
|
|
725
748
|
|
|
726
749
|
def build_image_generate_payload(**values: Any) -> dict[str, Any]:
|
|
@@ -735,22 +758,25 @@ def build_image_generate_payload(**values: Any) -> dict[str, Any]:
|
|
|
735
758
|
values["controls"] = json.loads(controls)
|
|
736
759
|
except json.JSONDecodeError as exc:
|
|
737
760
|
raise InputParsingError("Unable to parse Recraft controls JSON") from exc
|
|
738
|
-
|
|
739
|
-
|
|
761
|
+
return dump_image_payload(
|
|
762
|
+
_IMAGE_GENERATE_ADAPTER.validate_python(
|
|
763
|
+
{key: value for key, value in values.items() if value is not None}
|
|
764
|
+
)
|
|
740
765
|
)
|
|
741
|
-
return builder.model_validate(
|
|
742
|
-
{key: value for key, value in values.items() if value is not None}
|
|
743
|
-
).model_dump(mode="json", exclude_none=True)
|
|
744
766
|
|
|
745
767
|
|
|
746
768
|
def build_image_edit_payload(**values: Any) -> dict[str, Any]:
|
|
747
769
|
"""Validate image edit payloads and encode every local source image."""
|
|
748
770
|
|
|
749
771
|
images = values.pop("images")
|
|
750
|
-
return
|
|
751
|
-
|
|
752
|
-
|
|
753
|
-
|
|
772
|
+
return dump_image_payload(
|
|
773
|
+
_IMAGE_EDIT_ADAPTER.validate_python(
|
|
774
|
+
{
|
|
775
|
+
**{key: value for key, value in values.items() if value is not None},
|
|
776
|
+
"image_urls": [media_value(item) for item in images],
|
|
777
|
+
}
|
|
778
|
+
)
|
|
779
|
+
)
|
|
754
780
|
|
|
755
781
|
|
|
756
782
|
def build_image_vectorize_payload(
|
|
@@ -860,9 +886,27 @@ def build_lipsync_video_payload(**values: Any) -> dict[str, Any]:
|
|
|
860
886
|
|
|
861
887
|
video = values.pop("video")
|
|
862
888
|
audio = values.pop("audio")
|
|
863
|
-
|
|
889
|
+
payload = LipsyncVideoPayload(
|
|
864
890
|
**values, video_url=media_value(video), audio_url=media_value(audio)
|
|
865
|
-
)
|
|
891
|
+
)
|
|
892
|
+
return payload.model_dump(
|
|
893
|
+
exclude_none=True, exclude={"model"} if "model" not in values else set()
|
|
894
|
+
)
|
|
895
|
+
|
|
896
|
+
|
|
897
|
+
_LIPSYNC_IMAGE_BUILDERS: dict[str, type[BaseModel]] = {
|
|
898
|
+
"inftalk": LipsyncImagePayload,
|
|
899
|
+
"ltx23": LipsyncImagePayload,
|
|
900
|
+
"heygen-avatar-4": HeyGenImageLipsyncPayload,
|
|
901
|
+
"heygen-avatar-6": HeyGenAvatar6Payload,
|
|
902
|
+
"heygen-avatar-6-turbo": HeyGenAvatar6Payload,
|
|
903
|
+
}
|
|
904
|
+
|
|
905
|
+
|
|
906
|
+
_LIPSYNC_EXPRESSIVENESS_FIELDS = {
|
|
907
|
+
"heygen-avatar-6": "motion_expressiveness",
|
|
908
|
+
"heygen-avatar-6-turbo": "motion_expressiveness",
|
|
909
|
+
}
|
|
866
910
|
|
|
867
911
|
|
|
868
912
|
def build_lipsync_image_payload(**values: Any) -> dict[str, Any]:
|
|
@@ -870,9 +914,34 @@ def build_lipsync_image_payload(**values: Any) -> dict[str, Any]:
|
|
|
870
914
|
|
|
871
915
|
image = values.pop("image")
|
|
872
916
|
audio = values.pop("audio")
|
|
873
|
-
|
|
874
|
-
|
|
875
|
-
|
|
917
|
+
expressiveness = values.pop("expressiveness", None)
|
|
918
|
+
if expressiveness is not None:
|
|
919
|
+
field = _LIPSYNC_EXPRESSIVENESS_FIELDS.get(
|
|
920
|
+
values.get("model", "inftalk"), "expressiveness"
|
|
921
|
+
)
|
|
922
|
+
if values.get(field) is not None:
|
|
923
|
+
raise ValidationCLIError("Expressiveness was provided twice")
|
|
924
|
+
values[field] = expressiveness
|
|
925
|
+
builder = _LIPSYNC_IMAGE_BUILDERS.get(
|
|
926
|
+
values.get("model", "inftalk"), LipsyncImagePayload
|
|
927
|
+
)
|
|
928
|
+
return builder.model_validate(
|
|
929
|
+
{
|
|
930
|
+
**{key: value for key, value in values.items() if value is not None},
|
|
931
|
+
"image_url": media_value(image),
|
|
932
|
+
**({"audio_url": media_value(audio)} if audio is not None else {}),
|
|
933
|
+
}
|
|
934
|
+
).model_dump(mode="json", exclude_none=True)
|
|
935
|
+
|
|
936
|
+
|
|
937
|
+
def build_video_translate_payload(*, video: str, **values: Any) -> dict[str, Any]:
|
|
938
|
+
"""Validate translation inputs after encoding local video media."""
|
|
939
|
+
return VideoTranslatePayload.model_validate(
|
|
940
|
+
{
|
|
941
|
+
"video_url": media_value(video),
|
|
942
|
+
**values,
|
|
943
|
+
}
|
|
944
|
+
).model_dump(mode="json", exclude_none=True)
|
|
876
945
|
|
|
877
946
|
|
|
878
947
|
def build_media_payload(
|
|
@@ -1016,7 +1085,7 @@ class ApplicationServices:
|
|
|
1016
1085
|
*,
|
|
1017
1086
|
prompt: str,
|
|
1018
1087
|
model: str,
|
|
1019
|
-
image_size: str,
|
|
1088
|
+
image_size: str | None,
|
|
1020
1089
|
style: str | None,
|
|
1021
1090
|
image: str | None,
|
|
1022
1091
|
resolution: str | None,
|
|
@@ -1026,6 +1095,11 @@ class ApplicationServices:
|
|
|
1026
1095
|
lora_url: str | None = None,
|
|
1027
1096
|
controls: dict[str, Any] | str | None = None,
|
|
1028
1097
|
quality: str | None = None,
|
|
1098
|
+
complexity: str | None = None,
|
|
1099
|
+
output_format: str | None = None,
|
|
1100
|
+
output_compression: int | None = None,
|
|
1101
|
+
frame_count: int | None = None,
|
|
1102
|
+
frame_duration_ms: int | None = None,
|
|
1029
1103
|
) -> OperationResult:
|
|
1030
1104
|
payload = build_image_generate_payload(
|
|
1031
1105
|
prompt=prompt,
|
|
@@ -1039,14 +1113,17 @@ class ApplicationServices:
|
|
|
1039
1113
|
lora_url=lora_url,
|
|
1040
1114
|
controls=controls,
|
|
1041
1115
|
quality=quality,
|
|
1116
|
+
complexity=complexity,
|
|
1117
|
+
output_format=output_format,
|
|
1118
|
+
output_compression=output_compression,
|
|
1119
|
+
frame_count=frame_count,
|
|
1120
|
+
frame_duration_ms=frame_duration_ms,
|
|
1042
1121
|
)
|
|
1043
1122
|
return self._execute_async(
|
|
1044
1123
|
endpoint="/v1/image_generate",
|
|
1045
1124
|
payload=payload,
|
|
1046
1125
|
execution=execution,
|
|
1047
|
-
fallback_extension=(
|
|
1048
|
-
".svg" if model in _IMAGE_GENERATE_BUILDERS else ".png"
|
|
1049
|
-
),
|
|
1126
|
+
fallback_extension=image_fallback_extension(payload),
|
|
1050
1127
|
)
|
|
1051
1128
|
|
|
1052
1129
|
def execute_image_vectorize(
|
|
@@ -1077,8 +1154,13 @@ class ApplicationServices:
|
|
|
1077
1154
|
seed: int,
|
|
1078
1155
|
callback_url: str | None,
|
|
1079
1156
|
execution: ExecutionOptions,
|
|
1080
|
-
megapixel_ratio: float =
|
|
1157
|
+
megapixel_ratio: float | None = None,
|
|
1081
1158
|
quality: str | None = None,
|
|
1159
|
+
complexity: str | None = None,
|
|
1160
|
+
output_format: str | None = None,
|
|
1161
|
+
output_compression: int | None = None,
|
|
1162
|
+
frame_count: int | None = None,
|
|
1163
|
+
frame_duration_ms: int | None = None,
|
|
1082
1164
|
) -> OperationResult:
|
|
1083
1165
|
payload = build_image_edit_payload(
|
|
1084
1166
|
prompt=prompt,
|
|
@@ -1088,6 +1170,11 @@ class ApplicationServices:
|
|
|
1088
1170
|
resolution=resolution,
|
|
1089
1171
|
megapixel_ratio=megapixel_ratio,
|
|
1090
1172
|
quality=quality,
|
|
1173
|
+
complexity=complexity,
|
|
1174
|
+
output_format=output_format,
|
|
1175
|
+
output_compression=output_compression,
|
|
1176
|
+
frame_count=frame_count,
|
|
1177
|
+
frame_duration_ms=frame_duration_ms,
|
|
1091
1178
|
seed=seed,
|
|
1092
1179
|
callback_url=callback_url,
|
|
1093
1180
|
)
|
|
@@ -1095,7 +1182,7 @@ class ApplicationServices:
|
|
|
1095
1182
|
endpoint="/v1/image_edit",
|
|
1096
1183
|
payload=payload,
|
|
1097
1184
|
execution=execution,
|
|
1098
|
-
fallback_extension=
|
|
1185
|
+
fallback_extension=image_fallback_extension(payload),
|
|
1099
1186
|
)
|
|
1100
1187
|
|
|
1101
1188
|
def execute_image_upscale(
|
|
@@ -1282,8 +1369,10 @@ class ApplicationServices:
|
|
|
1282
1369
|
seed: int,
|
|
1283
1370
|
callback_url: str | None,
|
|
1284
1371
|
execution: ExecutionOptions,
|
|
1372
|
+
model: str | None = None,
|
|
1285
1373
|
) -> OperationResult:
|
|
1286
1374
|
payload = build_lipsync_video_payload(
|
|
1375
|
+
**({"model": model} if model is not None else {}),
|
|
1287
1376
|
video=video,
|
|
1288
1377
|
audio=audio,
|
|
1289
1378
|
resolution=resolution,
|
|
@@ -1302,15 +1391,23 @@ class ApplicationServices:
|
|
|
1302
1391
|
self,
|
|
1303
1392
|
*,
|
|
1304
1393
|
image: str,
|
|
1305
|
-
audio: str,
|
|
1394
|
+
audio: str | None,
|
|
1306
1395
|
model: str,
|
|
1307
|
-
resolution: str,
|
|
1396
|
+
resolution: str | None,
|
|
1308
1397
|
prompt: str | None,
|
|
1309
|
-
seed: int,
|
|
1398
|
+
seed: int | None,
|
|
1310
1399
|
callback_url: str | None,
|
|
1311
1400
|
execution: ExecutionOptions,
|
|
1401
|
+
text: str | None = None,
|
|
1402
|
+
voice_id: str | None = None,
|
|
1403
|
+
aspect_ratio: str | None = None,
|
|
1404
|
+
expressiveness: str | None = None,
|
|
1312
1405
|
) -> OperationResult:
|
|
1313
1406
|
payload = build_lipsync_image_payload(
|
|
1407
|
+
text=text,
|
|
1408
|
+
voice_id=voice_id,
|
|
1409
|
+
aspect_ratio=aspect_ratio,
|
|
1410
|
+
expressiveness=expressiveness,
|
|
1314
1411
|
image=image,
|
|
1315
1412
|
audio=audio,
|
|
1316
1413
|
model=model,
|
|
@@ -1326,6 +1423,56 @@ class ApplicationServices:
|
|
|
1326
1423
|
fallback_extension=".mp4",
|
|
1327
1424
|
)
|
|
1328
1425
|
|
|
1426
|
+
def execute_video_translate(
|
|
1427
|
+
self,
|
|
1428
|
+
*,
|
|
1429
|
+
video: str,
|
|
1430
|
+
output_language: str,
|
|
1431
|
+
model: str = "heygen-speed",
|
|
1432
|
+
callback_url: str | None = None,
|
|
1433
|
+
execution: ExecutionOptions,
|
|
1434
|
+
) -> OperationResult:
|
|
1435
|
+
payload = build_video_translate_payload(
|
|
1436
|
+
video=video,
|
|
1437
|
+
output_language=output_language,
|
|
1438
|
+
model=model,
|
|
1439
|
+
callback_url=callback_url,
|
|
1440
|
+
)
|
|
1441
|
+
return self._execute_async(
|
|
1442
|
+
endpoint="/v1/video_translate",
|
|
1443
|
+
payload=payload,
|
|
1444
|
+
execution=execution,
|
|
1445
|
+
fallback_extension=".mp4",
|
|
1446
|
+
)
|
|
1447
|
+
|
|
1448
|
+
def list_translation_languages(self) -> OperationResult:
|
|
1449
|
+
return self._operations.execute(
|
|
1450
|
+
OperationRequest(
|
|
1451
|
+
endpoint="/v1/video_translate/languages",
|
|
1452
|
+
method="GET",
|
|
1453
|
+
payload={},
|
|
1454
|
+
execution=ExecutionOptions(),
|
|
1455
|
+
)
|
|
1456
|
+
)
|
|
1457
|
+
|
|
1458
|
+
def list_heygen_voices(
|
|
1459
|
+
self,
|
|
1460
|
+
*,
|
|
1461
|
+
language: str | None = None,
|
|
1462
|
+
limit: int = 20,
|
|
1463
|
+
token: str | None = None,
|
|
1464
|
+
) -> OperationResult:
|
|
1465
|
+
return self._operations.execute(
|
|
1466
|
+
OperationRequest(
|
|
1467
|
+
endpoint="/v1/heygen/voices",
|
|
1468
|
+
method="GET",
|
|
1469
|
+
payload=HeyGenVoicesPayload(
|
|
1470
|
+
language=language, limit=limit, token=token
|
|
1471
|
+
).model_dump(exclude_none=True),
|
|
1472
|
+
execution=ExecutionOptions(),
|
|
1473
|
+
)
|
|
1474
|
+
)
|
|
1475
|
+
|
|
1329
1476
|
def execute_audio_transcribe(
|
|
1330
1477
|
self,
|
|
1331
1478
|
*,
|
|
@@ -1583,16 +1730,24 @@ class ApplicationServices:
|
|
|
1583
1730
|
payload = normalize_image_quality(payload)
|
|
1584
1731
|
except ValueError as exc:
|
|
1585
1732
|
raise ValidationCLIError(str(exc)) from exc
|
|
1586
|
-
|
|
1587
|
-
|
|
1588
|
-
|
|
1589
|
-
|
|
1590
|
-
|
|
1591
|
-
|
|
1592
|
-
|
|
1593
|
-
|
|
1594
|
-
|
|
1595
|
-
|
|
1733
|
+
image_adapter, models = (
|
|
1734
|
+
(_IMAGE_GENERATE_ADAPTER, ImageGenerateModel)
|
|
1735
|
+
if path == "/v1/image_generate"
|
|
1736
|
+
else (_IMAGE_EDIT_ADAPTER, ImageEditModel)
|
|
1737
|
+
)
|
|
1738
|
+
if "model" not in payload or payload.get("model") in get_args(models):
|
|
1739
|
+
image_adapter.validate_python(payload, extra="ignore")
|
|
1740
|
+
if path == "/v1/image_lipsync":
|
|
1741
|
+
builder = _LIPSYNC_IMAGE_BUILDERS.get(
|
|
1742
|
+
payload.get("model", "inftalk"), LipsyncImagePayload
|
|
1743
|
+
)
|
|
1744
|
+
builder.model_validate(payload, extra="ignore")
|
|
1745
|
+
if path == "/v1/video_lipsync":
|
|
1746
|
+
LipsyncVideoPayload.model_validate(payload, extra="ignore")
|
|
1747
|
+
if path == "/v1/video_translate":
|
|
1748
|
+
VideoTranslatePayload.model_validate(payload, extra="ignore")
|
|
1749
|
+
if path == "/v1/heygen/voices":
|
|
1750
|
+
HeyGenVoicesPayload.model_validate(payload)
|
|
1596
1751
|
if path == "/v1/video_generate" and payload.get("model") in get_args(
|
|
1597
1752
|
VideoGenerateModel
|
|
1598
1753
|
):
|
|
@@ -1645,9 +1800,8 @@ class ApplicationServices:
|
|
|
1645
1800
|
payload=payload,
|
|
1646
1801
|
execution=execution,
|
|
1647
1802
|
fallback_extension=(
|
|
1648
|
-
|
|
1649
|
-
if path
|
|
1650
|
-
and payload.get("model") in _IMAGE_GENERATE_BUILDERS
|
|
1803
|
+
image_fallback_extension(payload)
|
|
1804
|
+
if path in {"/v1/image_generate", "/v1/image_edit"}
|
|
1651
1805
|
else fallback_extension_for_endpoint(path)
|
|
1652
1806
|
),
|
|
1653
1807
|
)
|