everypixel-cli 0.5.2__tar.gz → 0.6.0__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.
Files changed (25) hide show
  1. {everypixel_cli-0.5.2 → everypixel_cli-0.6.0}/CHANGELOG.md +5 -0
  2. {everypixel_cli-0.5.2 → everypixel_cli-0.6.0}/PKG-INFO +43 -6
  3. {everypixel_cli-0.5.2 → everypixel_cli-0.6.0}/README.md +43 -6
  4. everypixel_cli-0.6.0/src/everypixel_cli/__init__.py +1 -0
  5. {everypixel_cli-0.5.2 → everypixel_cli-0.6.0}/src/everypixel_cli/application/services.py +178 -42
  6. {everypixel_cli-0.5.2 → everypixel_cli-0.6.0}/src/everypixel_cli/cli.py +113 -7
  7. {everypixel_cli-0.5.2 → everypixel_cli-0.6.0}/src/everypixel_cli/mcp_server.py +71 -5
  8. {everypixel_cli-0.5.2 → everypixel_cli-0.6.0}/src/everypixel_cli/openapi.py +17 -0
  9. {everypixel_cli-0.5.2 → everypixel_cli-0.6.0}/src/everypixel_cli/resources/openapi.json +3416 -1413
  10. {everypixel_cli-0.5.2 → everypixel_cli-0.6.0}/src/everypixel_cli/schemas.py +361 -207
  11. everypixel_cli-0.5.2/src/everypixel_cli/__init__.py +0 -1
  12. {everypixel_cli-0.5.2 → everypixel_cli-0.6.0}/.gitignore +0 -0
  13. {everypixel_cli-0.5.2 → everypixel_cli-0.6.0}/LICENSE +0 -0
  14. {everypixel_cli-0.5.2 → everypixel_cli-0.6.0}/pyproject.toml +0 -0
  15. {everypixel_cli-0.5.2 → everypixel_cli-0.6.0}/src/everypixel_cli/__main__.py +0 -0
  16. {everypixel_cli-0.5.2 → everypixel_cli-0.6.0}/src/everypixel_cli/application/__init__.py +0 -0
  17. {everypixel_cli-0.5.2 → everypixel_cli-0.6.0}/src/everypixel_cli/application/chat.py +0 -0
  18. {everypixel_cli-0.5.2 → everypixel_cli-0.6.0}/src/everypixel_cli/application/models.py +0 -0
  19. {everypixel_cli-0.5.2 → everypixel_cli-0.6.0}/src/everypixel_cli/application/serialization.py +0 -0
  20. {everypixel_cli-0.5.2 → everypixel_cli-0.6.0}/src/everypixel_cli/client.py +0 -0
  21. {everypixel_cli-0.5.2 → everypixel_cli-0.6.0}/src/everypixel_cli/config.py +0 -0
  22. {everypixel_cli-0.5.2 → everypixel_cli-0.6.0}/src/everypixel_cli/errors.py +0 -0
  23. {everypixel_cli-0.5.2 → everypixel_cli-0.6.0}/src/everypixel_cli/files.py +0 -0
  24. {everypixel_cli-0.5.2 → everypixel_cli-0.6.0}/src/everypixel_cli/output.py +0 -0
  25. {everypixel_cli-0.5.2 → everypixel_cli-0.6.0}/src/everypixel_cli/resources/__init__.py +0 -0
@@ -1,5 +1,10 @@
1
1
  # Changelog
2
2
 
3
+ ## [0.6.0] - 2026-11-16
4
+
5
+ - Add GPT Image 2.6 Sunburst with format, compression, and animation controls.
6
+ - Add HeyGen Avatar 4, video lipsync, translation, and voice/language catalogs.
7
+
3
8
  ## [0.5.2] - 2026-11-04
4
9
 
5
10
  - 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: 0.5.2
3
+ Version: 0.6.0
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` is supported only by `zimage`.
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` and `gpt-image-2.5-sunburst`. Both generation
204
- and editing accept `--quality low|medium|high` (default: `medium`); Sunburst also
205
- supports `xhigh` and `max`. Use `--size square` with `--resolution 3k`.
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-2.5-sunburst --quality high --download ./outputs
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),
@@ -314,6 +329,28 @@ 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-only 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
+ ```bash
340
+ everypixel lipsync heygen-voices --language English --limit 20
341
+
342
+ everypixel lipsync image --model heygen-avatar-4 --image ./portrait.png \
343
+ --text "Hello!" --voice-id VOICE_ID --download ./outputs
344
+
345
+ everypixel video translation-languages
346
+
347
+ everypixel video translate --video ./input.mp4 --output-language Spanish \
348
+ --model heygen-precision --download ./outputs
349
+ ```
350
+
351
+ Translation creates one task for one target language. Use `--estimate-cost`
352
+ with HeyGen lipsync or translation to request a quote.
353
+
317
354
  `--speaker`, `--style`, `--language`, and `--seed` are Qwen controls;
318
355
  `--character` applies only to Qwen voice design. ElevenLabs synthesis uses
319
356
  `--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` is supported only by `zimage`.
168
-
169
- GPT Image models are `gpt-image-2` and `gpt-image-2.5-sunburst`. Both generation
170
- and editing accept `--quality low|medium|high` (default: `medium`); Sunburst also
171
- supports `xhigh` and `max`. Use `--size square` with `--resolution 3k`.
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-2.5-sunburst --quality high --download ./outputs
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),
@@ -280,6 +295,28 @@ 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-only 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
+ ```bash
306
+ everypixel lipsync heygen-voices --language English --limit 20
307
+
308
+ everypixel lipsync image --model heygen-avatar-4 --image ./portrait.png \
309
+ --text "Hello!" --voice-id VOICE_ID --download ./outputs
310
+
311
+ everypixel video translation-languages
312
+
313
+ everypixel video translate --video ./input.mp4 --output-language Spanish \
314
+ --model heygen-precision --download ./outputs
315
+ ```
316
+
317
+ Translation creates one task for one target language. Use `--estimate-cost`
318
+ with HeyGen lipsync or translation to request a quote.
319
+
283
320
  `--speaker`, `--style`, `--language`, and `--seed` are Qwen controls;
284
321
  `--character` applies only to Qwen voice design. ElevenLabs synthesis uses
285
322
  `--voice-id` instead of preset controls. `tts-clone` and `tts-design` default
@@ -0,0 +1 @@
1
+ __version__ = "0.6.0"
@@ -42,14 +42,18 @@ from ..schemas import (
42
42
  TTSVoiceSavePayload,
43
43
  TTSVoicesPayload,
44
44
  TTSVoiceDeletePayload,
45
- GPT_IMAGE_MODEL_QUALITIES,
45
+ HeyGenImageLipsyncPayload,
46
+ HeyGenVoicesPayload,
47
+ VideoTranslatePayload,
48
+ RecraftVectorModel,
49
+ ImageEditModel,
50
+ ImageGenerateModel,
46
51
  ImageEditPayload,
47
52
  ImageGeneratePayload,
48
53
  ImageUpscalePayload,
49
54
  ImageVectorizePayload,
50
55
  LipsyncImagePayload,
51
56
  LipsyncVideoPayload,
52
- RecraftGeneratePayload,
53
57
  TaskResponse,
54
58
  VideoEditRequest,
55
59
  VideoGenerateRequest,
@@ -79,6 +83,7 @@ COST_ESTIMATE_ENDPOINTS = frozenset(
79
83
  "image_edit_colors",
80
84
  "video_lipsync",
81
85
  "image_lipsync",
86
+ "video_translate",
82
87
  "video_generate",
83
88
  "video_edit",
84
89
  "video_upscale",
@@ -717,10 +722,26 @@ def build_video_edit_payload(
717
722
  return builder(inputs)
718
723
 
719
724
 
720
- _IMAGE_GENERATE_BUILDERS: dict[str, type[BaseModel]] = {
721
- "recraftv4_1_vector": RecraftGeneratePayload,
722
- "recraftv4_1_pro_vector": RecraftGeneratePayload,
723
- }
725
+ _IMAGE_GENERATE_ADAPTER: TypeAdapter[ImageGeneratePayload] = TypeAdapter(
726
+ ImageGeneratePayload
727
+ )
728
+ _IMAGE_EDIT_ADAPTER: TypeAdapter[ImageEditPayload] = TypeAdapter(ImageEditPayload)
729
+
730
+
731
+ def dump_image_payload(payload: BaseModel) -> dict[str, Any]:
732
+ # GIF sizing is automatic; sending materialized size defaults is invalid.
733
+ exclude = (
734
+ {"image_size", "resolution"}
735
+ if getattr(payload, "output_format", None) == "gif"
736
+ else set()
737
+ )
738
+ return payload.model_dump(mode="json", exclude_none=True, exclude=exclude)
739
+
740
+
741
+ def image_fallback_extension(payload: Mapping[str, Any]) -> str:
742
+ if payload.get("model") in get_args(RecraftVectorModel):
743
+ return ".svg"
744
+ return "." + str(payload.get("output_format", "png"))
724
745
 
725
746
 
726
747
  def build_image_generate_payload(**values: Any) -> dict[str, Any]:
@@ -735,22 +756,25 @@ def build_image_generate_payload(**values: Any) -> dict[str, Any]:
735
756
  values["controls"] = json.loads(controls)
736
757
  except json.JSONDecodeError as exc:
737
758
  raise InputParsingError("Unable to parse Recraft controls JSON") from exc
738
- builder = _IMAGE_GENERATE_BUILDERS.get(
739
- values.get("model", "zimage"), ImageGeneratePayload
759
+ return dump_image_payload(
760
+ _IMAGE_GENERATE_ADAPTER.validate_python(
761
+ {key: value for key, value in values.items() if value is not None}
762
+ )
740
763
  )
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
764
 
745
765
 
746
766
  def build_image_edit_payload(**values: Any) -> dict[str, Any]:
747
767
  """Validate image edit payloads and encode every local source image."""
748
768
 
749
769
  images = values.pop("images")
750
- return ImageEditPayload(
751
- **{key: value for key, value in values.items() if value is not None},
752
- image_urls=[media_value(item) for item in images],
753
- ).model_dump(mode="json", exclude_none=True)
770
+ return dump_image_payload(
771
+ _IMAGE_EDIT_ADAPTER.validate_python(
772
+ {
773
+ **{key: value for key, value in values.items() if value is not None},
774
+ "image_urls": [media_value(item) for item in images],
775
+ }
776
+ )
777
+ )
754
778
 
755
779
 
756
780
  def build_image_vectorize_payload(
@@ -860,9 +884,19 @@ def build_lipsync_video_payload(**values: Any) -> dict[str, Any]:
860
884
 
861
885
  video = values.pop("video")
862
886
  audio = values.pop("audio")
863
- return LipsyncVideoPayload(
887
+ payload = LipsyncVideoPayload(
864
888
  **values, video_url=media_value(video), audio_url=media_value(audio)
865
- ).model_dump(exclude_none=True)
889
+ )
890
+ return payload.model_dump(
891
+ exclude_none=True, exclude={"model"} if "model" not in values else set()
892
+ )
893
+
894
+
895
+ _LIPSYNC_IMAGE_BUILDERS: dict[str, type[BaseModel]] = {
896
+ "inftalk": LipsyncImagePayload,
897
+ "ltx23": LipsyncImagePayload,
898
+ "heygen-avatar-4": HeyGenImageLipsyncPayload,
899
+ }
866
900
 
867
901
 
868
902
  def build_lipsync_image_payload(**values: Any) -> dict[str, Any]:
@@ -870,9 +904,26 @@ def build_lipsync_image_payload(**values: Any) -> dict[str, Any]:
870
904
 
871
905
  image = values.pop("image")
872
906
  audio = values.pop("audio")
873
- return LipsyncImagePayload(
874
- **values, image_url=media_value(image), audio_url=media_value(audio)
875
- ).model_dump(exclude_none=True)
907
+ builder = _LIPSYNC_IMAGE_BUILDERS.get(
908
+ values.get("model", "inftalk"), LipsyncImagePayload
909
+ )
910
+ return builder.model_validate(
911
+ {
912
+ **{key: value for key, value in values.items() if value is not None},
913
+ "image_url": media_value(image),
914
+ **({"audio_url": media_value(audio)} if audio is not None else {}),
915
+ }
916
+ ).model_dump(mode="json", exclude_none=True)
917
+
918
+
919
+ def build_video_translate_payload(*, video: str, **values: Any) -> dict[str, Any]:
920
+ """Validate translation inputs after encoding local video media."""
921
+ return VideoTranslatePayload.model_validate(
922
+ {
923
+ "video_url": media_value(video),
924
+ **values,
925
+ }
926
+ ).model_dump(mode="json", exclude_none=True)
876
927
 
877
928
 
878
929
  def build_media_payload(
@@ -1016,7 +1067,7 @@ class ApplicationServices:
1016
1067
  *,
1017
1068
  prompt: str,
1018
1069
  model: str,
1019
- image_size: str,
1070
+ image_size: str | None,
1020
1071
  style: str | None,
1021
1072
  image: str | None,
1022
1073
  resolution: str | None,
@@ -1026,6 +1077,11 @@ class ApplicationServices:
1026
1077
  lora_url: str | None = None,
1027
1078
  controls: dict[str, Any] | str | None = None,
1028
1079
  quality: str | None = None,
1080
+ complexity: str | None = None,
1081
+ output_format: str | None = None,
1082
+ output_compression: int | None = None,
1083
+ frame_count: int | None = None,
1084
+ frame_duration_ms: int | None = None,
1029
1085
  ) -> OperationResult:
1030
1086
  payload = build_image_generate_payload(
1031
1087
  prompt=prompt,
@@ -1039,14 +1095,17 @@ class ApplicationServices:
1039
1095
  lora_url=lora_url,
1040
1096
  controls=controls,
1041
1097
  quality=quality,
1098
+ complexity=complexity,
1099
+ output_format=output_format,
1100
+ output_compression=output_compression,
1101
+ frame_count=frame_count,
1102
+ frame_duration_ms=frame_duration_ms,
1042
1103
  )
1043
1104
  return self._execute_async(
1044
1105
  endpoint="/v1/image_generate",
1045
1106
  payload=payload,
1046
1107
  execution=execution,
1047
- fallback_extension=(
1048
- ".svg" if model in _IMAGE_GENERATE_BUILDERS else ".png"
1049
- ),
1108
+ fallback_extension=image_fallback_extension(payload),
1050
1109
  )
1051
1110
 
1052
1111
  def execute_image_vectorize(
@@ -1077,8 +1136,13 @@ class ApplicationServices:
1077
1136
  seed: int,
1078
1137
  callback_url: str | None,
1079
1138
  execution: ExecutionOptions,
1080
- megapixel_ratio: float = 1.0,
1139
+ megapixel_ratio: float | None = None,
1081
1140
  quality: str | None = None,
1141
+ complexity: str | None = None,
1142
+ output_format: str | None = None,
1143
+ output_compression: int | None = None,
1144
+ frame_count: int | None = None,
1145
+ frame_duration_ms: int | None = None,
1082
1146
  ) -> OperationResult:
1083
1147
  payload = build_image_edit_payload(
1084
1148
  prompt=prompt,
@@ -1088,6 +1152,11 @@ class ApplicationServices:
1088
1152
  resolution=resolution,
1089
1153
  megapixel_ratio=megapixel_ratio,
1090
1154
  quality=quality,
1155
+ complexity=complexity,
1156
+ output_format=output_format,
1157
+ output_compression=output_compression,
1158
+ frame_count=frame_count,
1159
+ frame_duration_ms=frame_duration_ms,
1091
1160
  seed=seed,
1092
1161
  callback_url=callback_url,
1093
1162
  )
@@ -1095,7 +1164,7 @@ class ApplicationServices:
1095
1164
  endpoint="/v1/image_edit",
1096
1165
  payload=payload,
1097
1166
  execution=execution,
1098
- fallback_extension=".png",
1167
+ fallback_extension=image_fallback_extension(payload),
1099
1168
  )
1100
1169
 
1101
1170
  def execute_image_upscale(
@@ -1282,8 +1351,10 @@ class ApplicationServices:
1282
1351
  seed: int,
1283
1352
  callback_url: str | None,
1284
1353
  execution: ExecutionOptions,
1354
+ model: str | None = None,
1285
1355
  ) -> OperationResult:
1286
1356
  payload = build_lipsync_video_payload(
1357
+ **({"model": model} if model is not None else {}),
1287
1358
  video=video,
1288
1359
  audio=audio,
1289
1360
  resolution=resolution,
@@ -1302,15 +1373,23 @@ class ApplicationServices:
1302
1373
  self,
1303
1374
  *,
1304
1375
  image: str,
1305
- audio: str,
1376
+ audio: str | None,
1306
1377
  model: str,
1307
- resolution: str,
1378
+ resolution: str | None,
1308
1379
  prompt: str | None,
1309
- seed: int,
1380
+ seed: int | None,
1310
1381
  callback_url: str | None,
1311
1382
  execution: ExecutionOptions,
1383
+ text: str | None = None,
1384
+ voice_id: str | None = None,
1385
+ aspect_ratio: str | None = None,
1386
+ expressiveness: str | None = None,
1312
1387
  ) -> OperationResult:
1313
1388
  payload = build_lipsync_image_payload(
1389
+ text=text,
1390
+ voice_id=voice_id,
1391
+ aspect_ratio=aspect_ratio,
1392
+ expressiveness=expressiveness,
1314
1393
  image=image,
1315
1394
  audio=audio,
1316
1395
  model=model,
@@ -1326,6 +1405,56 @@ class ApplicationServices:
1326
1405
  fallback_extension=".mp4",
1327
1406
  )
1328
1407
 
1408
+ def execute_video_translate(
1409
+ self,
1410
+ *,
1411
+ video: str,
1412
+ output_language: str,
1413
+ model: str = "heygen-speed",
1414
+ callback_url: str | None = None,
1415
+ execution: ExecutionOptions,
1416
+ ) -> OperationResult:
1417
+ payload = build_video_translate_payload(
1418
+ video=video,
1419
+ output_language=output_language,
1420
+ model=model,
1421
+ callback_url=callback_url,
1422
+ )
1423
+ return self._execute_async(
1424
+ endpoint="/v1/video_translate",
1425
+ payload=payload,
1426
+ execution=execution,
1427
+ fallback_extension=".mp4",
1428
+ )
1429
+
1430
+ def list_translation_languages(self) -> OperationResult:
1431
+ return self._operations.execute(
1432
+ OperationRequest(
1433
+ endpoint="/v1/video_translate/languages",
1434
+ method="GET",
1435
+ payload={},
1436
+ execution=ExecutionOptions(),
1437
+ )
1438
+ )
1439
+
1440
+ def list_heygen_voices(
1441
+ self,
1442
+ *,
1443
+ language: str | None = None,
1444
+ limit: int = 20,
1445
+ token: str | None = None,
1446
+ ) -> OperationResult:
1447
+ return self._operations.execute(
1448
+ OperationRequest(
1449
+ endpoint="/v1/heygen/voices",
1450
+ method="GET",
1451
+ payload=HeyGenVoicesPayload(
1452
+ language=language, limit=limit, token=token
1453
+ ).model_dump(exclude_none=True),
1454
+ execution=ExecutionOptions(),
1455
+ )
1456
+ )
1457
+
1329
1458
  def execute_audio_transcribe(
1330
1459
  self,
1331
1460
  *,
@@ -1583,16 +1712,24 @@ class ApplicationServices:
1583
1712
  payload = normalize_image_quality(payload)
1584
1713
  except ValueError as exc:
1585
1714
  raise ValidationCLIError(str(exc)) from exc
1586
- if (
1587
- isinstance(payload.get("model"), str)
1588
- and payload["model"] in GPT_IMAGE_MODEL_QUALITIES
1589
- ):
1590
- image_schema = (
1591
- ImageGeneratePayload
1592
- if path == "/v1/image_generate"
1593
- else ImageEditPayload
1594
- )
1595
- image_schema.model_validate(payload, extra="ignore")
1715
+ image_adapter, models = (
1716
+ (_IMAGE_GENERATE_ADAPTER, ImageGenerateModel)
1717
+ if path == "/v1/image_generate"
1718
+ else (_IMAGE_EDIT_ADAPTER, ImageEditModel)
1719
+ )
1720
+ if "model" not in payload or payload.get("model") in get_args(models):
1721
+ image_adapter.validate_python(payload, extra="ignore")
1722
+ if path == "/v1/image_lipsync":
1723
+ builder = _LIPSYNC_IMAGE_BUILDERS.get(
1724
+ payload.get("model", "inftalk"), LipsyncImagePayload
1725
+ )
1726
+ builder.model_validate(payload, extra="ignore")
1727
+ if path == "/v1/video_lipsync":
1728
+ LipsyncVideoPayload.model_validate(payload, extra="ignore")
1729
+ if path == "/v1/video_translate":
1730
+ VideoTranslatePayload.model_validate(payload, extra="ignore")
1731
+ if path == "/v1/heygen/voices":
1732
+ HeyGenVoicesPayload.model_validate(payload)
1596
1733
  if path == "/v1/video_generate" and payload.get("model") in get_args(
1597
1734
  VideoGenerateModel
1598
1735
  ):
@@ -1645,9 +1782,8 @@ class ApplicationServices:
1645
1782
  payload=payload,
1646
1783
  execution=execution,
1647
1784
  fallback_extension=(
1648
- ".svg"
1649
- if path == "/v1/image_generate"
1650
- and payload.get("model") in _IMAGE_GENERATE_BUILDERS
1785
+ image_fallback_extension(payload)
1786
+ if path in {"/v1/image_generate", "/v1/image_edit"}
1651
1787
  else fallback_extension_for_endpoint(path)
1652
1788
  ),
1653
1789
  )