visual-ai-assertions 0.11.0 → 0.13.0

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 (8) hide show
  1. package/README.md +35 -27
  2. package/dist/index.cjs +50 -5
  3. package/dist/index.cjs.map +1 -1
  4. package/dist/index.d.cts +8 -1
  5. package/dist/index.d.ts +8 -1
  6. package/dist/index.js +50 -5
  7. package/dist/index.js.map +1 -1
  8. package/package.json +1 -1
package/README.md CHANGED
@@ -205,7 +205,8 @@ const result = await ai.compare(before, after, {
205
205
  instructions: ["Ignore date/time differences"],
206
206
  });
207
207
 
208
- // With AI-generated diff image (supported only by gemini-3-flash-preview)
208
+ // With AI-generated diff image (supported by gemini-3-flash-preview and gemini-3.5-flash;
209
+ // only gemini-3-flash-preview auto-enables it — pass diffImage: true explicitly for 3.5-flash)
209
210
  const result = await ai.compare(before, after, {
210
211
  diffImage: true,
211
212
  });
@@ -490,12 +491,12 @@ const ai = visualAI({
490
491
 
491
492
  When omitted, each provider uses its default behavior. The `"xhigh"` level enables maximum reasoning depth.
492
493
 
493
- | Provider | Native Parameter | `"xhigh"` maps to |
494
- | ------------------ | ----------------------------------------------------- | -------------------- |
495
- | Anthropic Opus 4.7 | `thinking.type: "adaptive"` + `output_config.effort` | `effort: "xhigh"` |
496
- | Anthropic (other) | `thinking.type: "adaptive"` + `output_config.effort` | `effort: "max"` |
497
- | OpenAI | `reasoning.effort` (Responses API) | `effort: "xhigh"` |
498
- | Google | `thinkingConfig.thinkingBudget` (1024 / 8192 / 24576) | `24576` (max budget) |
494
+ | Provider | Native Parameter | `"xhigh"` maps to |
495
+ | ----------------------------------------- | ----------------------------------------------------- | -------------------- |
496
+ | Anthropic (Fable 5/Opus 4.8/4.7/Sonnet 5) | `thinking.type: "adaptive"` + `output_config.effort` | `effort: "xhigh"` |
497
+ | Anthropic (other) | `thinking.type: "adaptive"` + `output_config.effort` | `effort: "max"` |
498
+ | OpenAI | `reasoning.effort` (Responses API) | `effort: "xhigh"` |
499
+ | Google | `thinkingConfig.thinkingBudget` (1024 / 8192 / 24576) | `24576` (max budget) |
499
500
 
500
501
  ## Supported Models
501
502
 
@@ -503,32 +504,39 @@ All listed models support image/vision input. Pass any model ID to the `model` c
503
504
 
504
505
  ### Anthropic
505
506
 
506
- | Model | Model ID | Input $/MTok | Output $/MTok | Notes |
507
- | ----------------- | ------------------- | ------------ | ------------- | ------------------------------------------ |
508
- | Claude Opus 4.7 | `claude-opus-4-7` | $5 | $25 | Most capable; supports `xhigh` effort tier |
509
- | Claude Opus 4.6 | `claude-opus-4-6` | $5 | $25 | Previous flagship, 128K max output |
510
- | Claude Sonnet 4.6 | `claude-sonnet-4-6` | $3 | $15 | **Default** — best value |
511
- | Claude Haiku 4.5 | `claude-haiku-4-5` | $1 | $5 | Fastest, budget-friendly |
507
+ | Model | Model ID | Input $/MTok | Output $/MTok | Notes |
508
+ | ----------------- | ------------------- | ------------ | ------------- | ------------------------------------------- |
509
+ | Claude Fable 5 | `claude-fable-5` | $10 | $50 | Most capable; long-horizon agentic work |
510
+ | Claude Opus 4.8 | `claude-opus-4-8` | $5 | $25 | Most capable Opus tier; supports `xhigh` |
511
+ | Claude Opus 4.7 | `claude-opus-4-7` | $5 | $25 | Previous Opus; supports `xhigh` effort tier |
512
+ | Claude Opus 4.6 | `claude-opus-4-6` | $5 | $25 | Previous flagship, 128K max output |
513
+ | Claude Sonnet 5 | `claude-sonnet-5` | $3 | $15 | Near-Opus quality on coding/agentic work |
514
+ | Claude Sonnet 4.6 | `claude-sonnet-4-6` | $3 | $15 | **Default** — best value |
515
+ | Claude Haiku 4.5 | `claude-haiku-4-5` | $1 | $5 | Fastest, budget-friendly |
512
516
 
513
517
  ### OpenAI
514
518
 
515
- | Model | Model ID | Input $/MTok | Output $/MTok | Notes |
516
- | ------------ | -------------- | ------------ | ------------- | ------------------------------ |
517
- | GPT-5.5 | `gpt-5.5` | $5 | $30 | Newest flagship, 1M context |
518
- | GPT-5.4 Pro | `gpt-5.4-pro` | $30 | $180 | Most capable, extended context |
519
- | GPT-5.4 | `gpt-5.4` | $2.50 | $15 | Best vision quality |
520
- | GPT-5.2 | `gpt-5.2` | $1.75 | $14 | Balanced quality and cost |
521
- | GPT-5.4 mini | `gpt-5.4-mini` | $0.75 | $4.50 | Fast and affordable |
522
- | GPT-5.4 nano | `gpt-5.4-nano` | $0.20 | $1.25 | Cheapest OpenAI option |
523
- | GPT-5 mini | `gpt-5-mini` | $0.25 | $2 | **Default** — fast and cheap |
519
+ | Model | Model ID | Input $/MTok | Output $/MTok | Notes |
520
+ | ------------- | --------------- | ------------ | ------------- | --------------------------------- |
521
+ | GPT-5.6 Sol | `gpt-5.6-sol` | $5 | $30 | Newest flagship, frontier tier |
522
+ | GPT-5.6 Terra | `gpt-5.6-terra` | $2.50 | $15 | Newest balanced, everyday tier |
523
+ | GPT-5.6 Luna | `gpt-5.6-luna` | $1 | $6 | Newest, fastest/cheapest tier |
524
+ | GPT-5.5 | `gpt-5.5` | $5 | $30 | Previous flagship, 1M context |
525
+ | GPT-5.4 Pro | `gpt-5.4-pro` | $30 | $180 | Most capable, extended context |
526
+ | GPT-5.4 | `gpt-5.4` | $2.50 | $15 | Best vision quality |
527
+ | GPT-5.2 | `gpt-5.2` | $1.75 | $14 | Balanced quality and cost |
528
+ | GPT-5.4 mini | `gpt-5.4-mini` | $0.75 | $4.50 | **Default** — fast and affordable |
529
+ | GPT-5.4 nano | `gpt-5.4-nano` | $0.20 | $1.25 | Cheapest older-generation option |
530
+ | GPT-5 mini | `gpt-5-mini` | $0.25 | $2 | Fast and cheap |
524
531
 
525
532
  ### Google
526
533
 
527
- | Model | Model ID | Input $/MTok | Output $/MTok | Notes |
528
- | --------------------- | ------------------------------- | ------------ | ------------- | --------------------------------- |
529
- | Gemini 3.1 Pro | `gemini-3.1-pro-preview` | $2 | $12 | Preview — most advanced reasoning |
530
- | Gemini 3.1 Flash Lite | `gemini-3.1-flash-lite-preview` | $0.25 | $1.50 | Preview — lightweight and cheap |
531
- | Gemini 3 Flash | `gemini-3-flash-preview` | $0.50 | $3 | **Default** — fast and capable |
534
+ | Model | Model ID | Input $/MTok | Output $/MTok | Notes |
535
+ | --------------------- | ------------------------ | ------------ | ------------- | --------------------------------- |
536
+ | Gemini 3.5 Flash | `gemini-3.5-flash` | $1.50 | $9 | Strongest agentic & coding model |
537
+ | Gemini 3.1 Pro | `gemini-3.1-pro-preview` | $2 | $12 | Preview — most advanced reasoning |
538
+ | Gemini 3.1 Flash Lite | `gemini-3.1-flash-lite` | $0.25 | $1.50 | GA — lightweight and cheap |
539
+ | Gemini 3 Flash | `gemini-3-flash-preview` | $0.50 | $3 | **Default** — fast and capable |
532
540
 
533
541
  ## License
534
542
 
package/dist/index.cjs CHANGED
@@ -80,12 +80,18 @@ var Provider = {
80
80
  };
81
81
  var Model = {
82
82
  Anthropic: {
83
+ FABLE_5: "claude-fable-5",
84
+ OPUS_4_8: "claude-opus-4-8",
83
85
  OPUS_4_7: "claude-opus-4-7",
84
86
  OPUS_4_6: "claude-opus-4-6",
87
+ SONNET_5: "claude-sonnet-5",
85
88
  SONNET_4_6: "claude-sonnet-4-6",
86
89
  HAIKU_4_5: "claude-haiku-4-5"
87
90
  },
88
91
  OpenAI: {
92
+ GPT_5_6_SOL: "gpt-5.6-sol",
93
+ GPT_5_6_TERRA: "gpt-5.6-terra",
94
+ GPT_5_6_LUNA: "gpt-5.6-luna",
89
95
  GPT_5_5: "gpt-5.5",
90
96
  GPT_5_4: "gpt-5.4",
91
97
  GPT_5_4_PRO: "gpt-5.4-pro",
@@ -95,8 +101,9 @@ var Model = {
95
101
  GPT_5_MINI: "gpt-5-mini"
96
102
  },
97
103
  Google: {
104
+ GEMINI_3_5_FLASH: "gemini-3.5-flash",
98
105
  GEMINI_3_1_PRO_PREVIEW: "gemini-3.1-pro-preview",
99
- GEMINI_3_1_FLASH_LITE_PREVIEW: "gemini-3.1-flash-lite-preview",
106
+ GEMINI_3_1_FLASH_LITE: "gemini-3.1-flash-lite",
100
107
  GEMINI_3_FLASH_PREVIEW: "gemini-3-flash-preview"
101
108
  }
102
109
  };
@@ -572,9 +579,15 @@ function parseRetryAfter(value) {
572
579
  }
573
580
 
574
581
  // src/providers/anthropic.ts
582
+ var XHIGH_CAPABLE_MODELS = /* @__PURE__ */ new Set([
583
+ Model.Anthropic.FABLE_5,
584
+ Model.Anthropic.OPUS_4_8,
585
+ Model.Anthropic.OPUS_4_7,
586
+ Model.Anthropic.SONNET_5
587
+ ]);
575
588
  function mapEffort(level, model) {
576
589
  if (level !== "xhigh") return level;
577
- return model === Model.Anthropic.OPUS_4_7 ? "xhigh" : "max";
590
+ return XHIGH_CAPABLE_MODELS.has(model) ? "xhigh" : "max";
578
591
  }
579
592
  var AnthropicDriver = class {
580
593
  client;
@@ -957,6 +970,18 @@ function resolveConfig(config) {
957
970
  // src/core/pricing.ts
958
971
  var PER_MILLION = 1e6;
959
972
  var PRICING_TABLE = {
973
+ [`${Provider.ANTHROPIC}:${Model.Anthropic.FABLE_5}`]: {
974
+ inputPricePerToken: 10 / PER_MILLION,
975
+ outputPricePerToken: 50 / PER_MILLION
976
+ },
977
+ [`${Provider.ANTHROPIC}:${Model.Anthropic.OPUS_4_8}`]: {
978
+ inputPricePerToken: 5 / PER_MILLION,
979
+ outputPricePerToken: 25 / PER_MILLION
980
+ },
981
+ [`${Provider.ANTHROPIC}:${Model.Anthropic.SONNET_5}`]: {
982
+ inputPricePerToken: 3 / PER_MILLION,
983
+ outputPricePerToken: 15 / PER_MILLION
984
+ },
960
985
  [`${Provider.ANTHROPIC}:${Model.Anthropic.OPUS_4_7}`]: {
961
986
  inputPricePerToken: 5 / PER_MILLION,
962
987
  outputPricePerToken: 25 / PER_MILLION
@@ -973,6 +998,18 @@ var PRICING_TABLE = {
973
998
  inputPricePerToken: 1 / PER_MILLION,
974
999
  outputPricePerToken: 5 / PER_MILLION
975
1000
  },
1001
+ [`${Provider.OPENAI}:${Model.OpenAI.GPT_5_6_SOL}`]: {
1002
+ inputPricePerToken: 5 / PER_MILLION,
1003
+ outputPricePerToken: 30 / PER_MILLION
1004
+ },
1005
+ [`${Provider.OPENAI}:${Model.OpenAI.GPT_5_6_TERRA}`]: {
1006
+ inputPricePerToken: 2.5 / PER_MILLION,
1007
+ outputPricePerToken: 15 / PER_MILLION
1008
+ },
1009
+ [`${Provider.OPENAI}:${Model.OpenAI.GPT_5_6_LUNA}`]: {
1010
+ inputPricePerToken: 1 / PER_MILLION,
1011
+ outputPricePerToken: 6 / PER_MILLION
1012
+ },
976
1013
  [`${Provider.OPENAI}:${Model.OpenAI.GPT_5_5}`]: {
977
1014
  inputPricePerToken: 5 / PER_MILLION,
978
1015
  outputPricePerToken: 30 / PER_MILLION
@@ -1001,11 +1038,15 @@ var PRICING_TABLE = {
1001
1038
  inputPricePerToken: 0.25 / PER_MILLION,
1002
1039
  outputPricePerToken: 2 / PER_MILLION
1003
1040
  },
1041
+ [`${Provider.GOOGLE}:${Model.Google.GEMINI_3_5_FLASH}`]: {
1042
+ inputPricePerToken: 1.5 / PER_MILLION,
1043
+ outputPricePerToken: 9 / PER_MILLION
1044
+ },
1004
1045
  [`${Provider.GOOGLE}:${Model.Google.GEMINI_3_1_PRO_PREVIEW}`]: {
1005
1046
  inputPricePerToken: 2 / PER_MILLION,
1006
1047
  outputPricePerToken: 12 / PER_MILLION
1007
1048
  },
1008
- [`${Provider.GOOGLE}:${Model.Google.GEMINI_3_1_FLASH_LITE_PREVIEW}`]: {
1049
+ [`${Provider.GOOGLE}:${Model.Google.GEMINI_3_1_FLASH_LITE}`]: {
1009
1050
  inputPricePerToken: 0.25 / PER_MILLION,
1010
1051
  outputPricePerToken: 1.5 / PER_MILLION
1011
1052
  },
@@ -1087,15 +1128,19 @@ async function timedSendMessage(driver, images, prompt, options) {
1087
1128
 
1088
1129
  // src/core/diff.ts
1089
1130
  var import_sharp = __toESM(require("sharp"), 1);
1131
+ var DIFF_ALLOWED_MODELS = /* @__PURE__ */ new Set([
1132
+ Model.Google.GEMINI_3_FLASH_PREVIEW,
1133
+ Model.Google.GEMINI_3_5_FLASH
1134
+ ]);
1090
1135
  async function generateAiDiff(imgA, imgB, model, driver) {
1091
1136
  if (!driver.generateImage) {
1092
1137
  throw new VisualAIConfigError(
1093
1138
  "AI-generated diff images require a provider that supports image generation. Currently only the Google (Gemini) provider supports this."
1094
1139
  );
1095
1140
  }
1096
- if (model !== Model.Google.GEMINI_3_FLASH_PREVIEW) {
1141
+ if (!DIFF_ALLOWED_MODELS.has(model)) {
1097
1142
  throw new VisualAIConfigError(
1098
- "Annotated diff images are only supported when visualAI is configured with the Google model gemini-3-flash-preview."
1143
+ `Annotated diff images are only supported with these Google models: ${[...DIFF_ALLOWED_MODELS].join(", ")}.`
1099
1144
  );
1100
1145
  }
1101
1146
  const response = await driver.generateImage([imgA, imgB], buildAiDiffPrompt(), {